auto: the WYOMING analysis of 6-16 November and its labels; three old analyses removed (LENZSPITZE 1 Nov, NORDEND 4-7 and 5 Nov)

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
maximeandClaude Opus 5.5 committed 2026-11-17 12:25:48 +01:00
1 parent 4fd623b3e1
commit 528336f9e5
51 files changed
+433 -657

No files matched your search

+18
View File
@@ -0,0 +1,18 @@
{"id": "WYOMING-1", "start": "2026-11-06T09:15Z", "end": "2026-11-07T07:30Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. The housing temperature rises from late morning with the outside air (MeteoSwiss Genève-Cointrin: a grey, mild day) and the pose follows it (scale up, pan down); in the raw image the nearest prism, that of the reference RD and of its twin target_01, drifts sideways against all the other prisms, which hardly move. The compensation carries this drift into the targets between RD and the far references target_12 and target_13: the corr_x of target_14, target_15 and target_02 go down clearly, those of the far targets a little, target_01 hardly moves since it is the prism of RD. The curves stay shifted in the evening, then come back during the night as the air and the housing cool, until the turning point of the next morning. Deduced cause: the shift follows the housing temperature within the hour and comes back with it, the housing following the outside air and not the sun; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. This is not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-2", "start": "2026-11-07T07:30Z", "end": "2026-11-08T07:00Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. The housing temperature rises strongly in the morning with the outside air (MeteoSwiss Genève-Cointrin: a sunny spell around midday) and the pose follows it (scale up, pan down): the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and the compensation passes it into the corr_x of target_14, target_15 and target_02, which go down clearly, the far targets a little. From the afternoon the air and the housing cool steadily through a clear, calm night close to freezing; the pose and the curves go back past their level and keep going the other way: by the end of the night target_14, target_15 and target_02 are clearly above their level. The abrupt shift of the morning of the 8th is labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it at equal temperature; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-3", "start": "2026-11-08T07:00Z", "end": "2026-11-09T09:45Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. After the cold night the housing temperature rises again quickly in the morning with the outside air (MeteoSwiss Genève-Cointrin: sunny morning) and stays high until the next morning; the pose follows it (scale up, pan down), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and the compensation passes this drift into the corr_x of target_14, target_15 and target_02, which go clearly below their level, the far targets a little. In the evening and the mild night the curves come back only partly, then stay still until the start of the next warming. The abrupt shift of the morning of the 8th is labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-4", "start": "2026-11-09T09:45Z", "end": "2026-11-10T04:45Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. The housing temperature rises again from mid-morning with the outside air, the pose follows it (scale up, pan down) and the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms; through the compensation the corr_x of target_14, target_15 and target_02 go down clearly, the far targets a little. The housing stays warm in the afternoon and evening, then cools a little during the night and the curves come back part of the way, until the turning point of the next morning. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-5", "start": "2026-11-10T04:45Z", "end": "2026-11-13T09:30Z", "family": "Thermal displacement", "description": "Thermal movement over several days, poorly compensated. In a mild, overcast and at times rainy spell (MeteoSwiss Genève-Cointrin), the housing temperature rises on the 10th, stays high all night, rises again on the 11th to the warmest of the range, comes down in the evening and stays flat all the 12th; then the bise sets in during the night to the 13th, with a cold dry wind, and the housing cools quickly. The pose follows it (scale and pan), and the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms; through the compensation the corr_x of target_14, target_15 and target_02 go down strongly until the afternoon of the 11th, then come back, stay still on the 12th, and go past their level in the cold morning of the 13th, where they are clearly above it. The corr_y of the far targets follow the temperature too, up while warm, down in the cold morning of the 13th, target_03 and target_04 more than the others, with their error_value rising. The gap of the 10th, the occultation of target_14 and the two jumps of the 11th are labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it at equal temperature; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-6", "start": "2026-11-13T09:30Z", "end": "2026-11-14T07:00Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. Under the bise (MeteoSwiss Genève-Cointrin: cold dry wind from the north-east, some sun around midday), the housing temperature warms up a little in the late morning, then drops again in the evening and all night to the coldest of the range. The pose follows it (scale down, pan up), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and the compensation carries it into the corr_x of target_14, target_15 and target_02, which come back part of the way in the afternoon then rise strongly above their level all night; the corr_y of the far targets target_03 to target_11 stay below their level. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-7", "start": "2026-11-14T07:00Z", "end": "2026-11-15T04:15Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. After the cold night the housing temperature rises in the morning with the outside air and the sun (MeteoSwiss Genève-Cointrin), stays moderate all day, then cools in the evening and the night as the wind drops and light rain begins. The pose follows it (scale and pan), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and through the compensation the corr_x of target_14, target_15 and target_02 come back part of the way toward their level in the morning, then rise again above it in the night; the corr_y of the far targets stay below their level in the day and come back in the evening. The abrupt shift of the morning and the one-measurement jump of the evening are labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-8", "start": "2026-11-15T04:15Z", "end": "2026-11-16T06:15Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. After the night rain, a mild south-west wind sets in (MeteoSwiss Genève-Cointrin) and the housing temperature rises clearly from the morning to the afternoon, then cools slowly in the evening and the night. The pose follows it (scale up, pan down), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and through the compensation the corr_x of target_14, target_15 and target_02 go from above their level to clearly below it in the afternoon, then come back only part of the way during the night; the corr_y of the far targets come back to their level. At the end of the range the curves are still a little shifted and the camera pose is no longer exported, while the measurements go on. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-9", "start": "2026-11-07T04:45Z", "end": "2026-11-07T05:15Z", "family": "Displacement", "description": "One-measurement jump of the far targets target_03 to target_11, in corr_x and a little in corr_y, then back: in the raw image these targets do not move, while the far references target_12 and target_13 jump alone for this measurement, without their error_value rising and without rejection; the compensation passes their jump into the targets close to them in the image, target_14, target_15 and target_02 hardly move. No image at this instant. The air is cold and very humid, calm (MeteoSwiss Genève-Cointrin). Cause not established (a disturbance of the long line of sight to the far references over the river is possible). This is not a movement of the targets: points to ignore.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-10", "start": "2026-11-07T13:00Z", "end": "2026-11-07T13:30Z", "family": "Obstruction", "description": "target_03 and target_04, on neighbouring piers, are rejected for one measurement, the point of target_03 having jumped far away, then back at the next; the other targets and the references do not move. Cause deduced from the signature of a passage in front of these two targets for the length of one measurement; no image at this instant. Points missing for one measurement: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-11", "start": "2026-11-08T06:45Z", "end": "2026-11-08T08:30Z", "family": "Displacement", "description": "Sudden shift and return, at dawn after the coldest clear night: for three measurements all the far prisms, references target_12 and target_13 included, shift sideways together in the raw image, while the nearest prism, that of the reference RD and target_01, does not follow; the compensation cannot represent it and the corr_x of target_14, target_15, target_02 and target_01 drop sharply, the corr_y of the far targets target_03 to target_09 rise, the error_value rises on several of them without any rejection; everything is back within the hour, as the sun comes out (MeteoSwiss Genève-Cointrin). The 07:00 image shows a haze over the river and the sun on the facades, nothing in front of the targets. Cause not established: a bending of the long lines of sight through the cold air over the river at sunrise, or a fast movement of the camera as its housing warms up suddenly, are both possible; no image falls during the shift. Points not to be trusted during this hour; not a movement of the structure.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-12", "start": "2026-11-10T08:15Z", "end": "2026-11-10T09:00Z", "family": "No data", "description": "Two measurements missing for all sensors. ALASKA and NADELHORN miss the same two measurements: a common cause outside the station (transmission or reception of the data), not a station failure; it is not established. The station runs: a synchronization falls in the gap, the voltage is normal, the images of the hour before and after exist (one image per hour). Nothing to do on the station.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-13", "start": "2026-11-11T07:00Z", "end": "2026-11-11T07:30Z", "family": "Displacement", "description": "One-measurement jump of the corr_y of the far targets target_03 to target_11, then back: in the raw image these targets do not move, while the far references target_12 and target_13 jump alone, their error_value clearly up but far from the threshold, that of target_03 and target_11 a little up too; the compensation passes the jump of these references into the targets close to them in the image. No image at this instant, nothing in front on the images of the hour before and after. Cause not established: the rain begins at that hour (MeteoSwiss Genève-Cointrin) and could veil the far references for a moment, a passage partly in front of the far end of the bridge is possible too. Not a movement: points to ignore.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-14", "start": "2026-11-11T09:30Z", "end": "2026-11-11T14:15Z", "family": "Obstruction", "description": "target_14 alone: its error_value rises gradually from mid-morning, peaks around noon and falls back in the early afternoon, with a dip in between; its corr_x goes down and its corr_y down too, more than its neighbour target_15, then both come back with the error_value. In the raw image only target_14 slides, the other prisms do not. On the 12:00 thumbnail, part of the spot of target_14 is darkened, the rest intact; before and after the spot is whole. The episode covers the rain of the morning (MeteoSwiss Genève-Cointrin) and ends soon after it stops: water on the prism, a drop or a drip from the deck, is likely, though not seen as such. Ignore the curve of target_14 over the range; not a movement.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-15", "start": "2026-11-11T14:30Z", "end": "2026-11-11T15:00Z", "family": "Obstruction", "description": "For one measurement the far references target_12 and target_13 and the far targets target_07 to target_11 are rejected, their points jumping in all directions, the compensation has no reference left and the camera pose is missing; the corr of target_02 to target_06, kept, jump strongly and those of target_01 and target_14 a little: the points of this measurement make no sense. All back at the next measurement. The raw positions of the nearer prisms hardly move, which rules out a shake of the camera by the gusts rising at that hour (MeteoSwiss Genève-Cointrin); everything beyond a given pier is lost at once. Cause deduced from the signature of a passage across the line of sight, in front of the far end of the bridge, for the length of one measurement; no image at this instant, nothing in front on those of 14:00 and 15:00. Points to ignore: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-16", "start": "2026-11-14T07:00Z", "end": "2026-11-14T08:30Z", "family": "Displacement", "description": "Sudden shift and return, at dawn after the coldest night of the bise, the same signature as on the 8th: for three measurements all the far prisms, references target_12 and target_13 included, shift sideways together in the raw image, while the nearest prism, that of the reference RD and target_01, does not follow; the corr_x of target_14, target_15, target_02 and target_01 drop sharply toward their level, then rise again part of the way when the far prisms come back, as the sun comes out (MeteoSwiss Genève-Cointrin). The 07:00 image shows the sun on the facades, nothing in front of the targets. Cause not established: a bending of the long lines of sight through the cold air over the river at sunrise, or a fast movement of the camera as its housing warms up suddenly, are both possible; no image falls during the shift. Points not to be trusted during this hour; not a movement of the structure.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-17", "start": "2026-11-14T20:45Z", "end": "2026-11-14T21:15Z", "family": "Displacement", "description": "One-measurement jump of the corr_y of the targets target_02 to target_11, with a smaller one in corr_x, then back: in the raw image these targets do not move, while the far references target_12 and target_13 jump alone for this measurement, without their error_value rising; the compensation passes their jump into the targets. No image at this instant. A cold, calm night under clearing bise (MeteoSwiss Genève-Cointrin). Cause not established (a disturbance of the long line of sight to the far references over the river is possible). Not a movement: points to ignore.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
{"id": "WYOMING-18", "start": "2026-11-15T14:00Z", "end": "2026-11-15T14:30Z", "family": "Obstruction", "description": "target_03 is rejected for one measurement, its point jumping far away, and the error_value of target_04 and target_15 rises at the same instant, the corr_x of target_04 jumping a little; everything is back at the next measurement and the references do not move. Cause deduced from the signature of a passage for the length of one measurement; no image at this instant. Point missing for one measurement: nothing to do.", "analysis": "WYOMING_2026-11-06T0805_2026-11-16T0627", "review": "2026-11-17"}
@@ -1,61 +0,0 @@
# Automatic labels: LENZSPITZE, 2026-11-01T10:30Z → 2026-11-01T11:40Z (UTC)
Range chosen by the user in the Data page; their hint: "Around 11:00, the track targets move a little; a train?". No label of `labeling/auto/LENZSPITZE.jsonl` touches the range (the last one, LENZSPITZE-64, ends on 01.11 at 00:00). 5 min rate, plus one measurement at 11:01 (these off-rate measurements, at minute 01, often return on this station); 16 instants, no missing step, no rejection, the 9 references retained everywhere, one camera pose at each instant (`inventory.py`).
## 1. My labels
### 2026-11-01T10:30Z → 2026-11-01T11:40Z
- **Map phenomenon**: The camera moves through a thermal effect
- **Family**: Thermal displacement. It is always labeled (D-117). Here the pose follows the warming of the housing and the `corr_y` come back with it in the afternoon. It is not Displacement: the return is seen over the day, and the same cycle returns every sunny day on the station.
- **Measured**: `metrics.py` from 31.10 to 02.11: housing temperature (`temperature_deg_c`, the only one filled) at 26.5 °C at 10h, 27.3 °C at 11h and 29.7 °C at 12h, higher at 13h, then dropping; r(cam_pan, temperature) = −0.81 over 49 h. `overview.py` over the day: start of the pose around 06:50 (scale and tilt), then a second phase from 11:05 (tilt), then 11:15 (pan dropping, rotation rising), return around 14:15-15:00. In the batch: tilt from 0.0139° to 0.0074° between 11:00 and 11:05, pan from 0.0116° to 0.0047° between 11:10 and 11:15, rotation from 0.002° to 0.013° between 11:15 and 11:40; scale at 0.99968-0.99973. Raw positions: `x` and `y` of all references and of the median of the targets move together (−0.5 px in `x` at 11:15). `summary.py --step 30min` from 03:00 to 16:00: the `corr_y` of 1018, 1019, 1016, 1008 and 1020 drops at 07:00, stays low until 10:30, then rises from 11:00 toward its night level, reached around 13:00. `--rigid` with base 10:30: the targets move as a block (`tx`, `ty`), except the own deviations described under the label of 8012. Bounds: the cycle touches both edges of the batch, with no neighboring label, hence 10:30 and 11:40 (rules, section 2).
- **Images**: 10:01, 11:00 and 12:01: full sun on the track, sharp shadows.
- **Confidence**: medium. The shape of the cycle is sure; the tilt step at 11:05 and that of the pan at 11:15 are more abrupt than an ordinary cycle and fall at the same instants as the train and as 8012. I read them as the cycle, because the pose keeps them and comes back in the afternoon.
### 2026-11-01T10:55Z → 2026-11-01T11:05Z
- **Map phenomenon**: A passage hides the targets for the time of one measurement (variant: the flux stays whole, the `error_value` rises and a second pattern appears)
- **Family**: Obstruction. The passage is seen on the image, and it touched a group of sensors, not the whole scene through a diffuse nuisance.
- **Measured**: at 11:00, the `error_value` of 35 sensors out of 41 exceeds their median by more than 5 MAD, references included (8004 the highest, under its threshold); no rejection (`overview.py --min 1`). `quality.py --table` at 1 min, with the calm reference of 30-31.10: at 11:00, median sharpness at 0.62 (0.26 to 0.48 on 8010, 8004, 8003, 1016, 1012, 101B, 1014; 0.86 to 0.89 on 8014 and 8017) and median `dominance` at 2.2 (up to 6 to 7.7 on 8010, 8003, 8004, 1016, 1014, 1012); the flux stays whole (median 1.08), except 1006 at 0.34 at 11:00 and 0.43 at 11:01, back to 1.03 at 11:05. At 11:01, the image is even less sharp (0.83); at 11:05, it is back. `ranges.py --rolling 3 --instants`: 15 sensors off level at 11:00 (`corr_y` of 1022 at −10 noises, 202H at +6, 1009, 1013, 8019 at −5, 1001, 1005, 1020, 8010, 102H, 1017; `corr_x` of 101B, 1012, 1013, 202H, 8019); 5 at 11:01 (`corr_x` of 1008, 1013, 1012, 8013, `corr_y` of 1020); these jumps come back at 11:05 (`summary.py --step 1min`, `corr_y` of 1022, 202H, 1005, 1001 back from 11:01 or 11:05). The step of the `corr_y` of 1018, 1019, 1016, 1008 and 1020 that starts at the same instant and stays belongs to the return of the thermal cycle.
- **Images**: on the image of 11:00, a train passes on the track to the left of the instrumented track, in transparency and motion blur. On the fixed-frame thumbnails, it covers 8003 and 8010 and passes at the bottom of the references 8011, 8012 and 8013; it is absent at 10:01 and at 12:01 (`thumbnails.py --zone 1300,750,1000,550`, `thumbnails.py --fixed` on 8012, 8011, 8013, 8010, 8003, 1017).
- **Confidence**: strong. The train is seen at the minute of the measurement, and the image quality gives its signature on the sensors it covers.
### 2026-11-01T11:05Z → 2026-11-01T11:40Z
- **Map phenomenon**: close to "The sun in the field or behind a spot" (variant: the background of a reference in the sun makes a small common jump of several targets, under the threshold); not seen here
- **Family**: Visibility. It is the light on a reference that changes, with nothing in front. According to the boundaries, a reference that moves away from the others and from the targets, with a sharp edge, is Visibility and not Thermal displacement. It is not Displacement: the signature (flux, `dominance`) is that of a spot whose lighting changes, and it returns on other days at the same hour.
- **Measured**: `capteur_jours.py 8012` (my script): flux of 8012 from 21,000 to 31,000 between 11:05 and 11:15, `dominance` from 0.59 to 0.34, raw `x` from 2043.51 to 2043.36 px, `corr_x` from 0.01 to 0.33, then everything stays so until 11:45. `overview.py --rigid 2026-11-01T10:30Z`: the residual `dx` of 8012 goes from 0.07 (11:10) to 0.34 px (11:15) and 0.41 px (11:20), to slowly decrease down to 0.16 px at 12:30; the other references stay under 0.1 px. `sigma0` rises from 0.0014 to 0.0030 at 11:15 and stays high. The residual of the targets 8003 (−0.08 px), 8010 (+0.11 px) and 1017 (+0.10 px) changes at the same instant. `corr_x` (residuals of `overview.py`): first deviation at 11:10 (8003 at +8 noises, 8004 at +6), then step at 11:15. 8003 drops by about −0.12, 1011 and 1001 by about −0.08, 1003, 1006, 1021, 201B, 7013H and 8004 by −0.05 to −0.08. 8010 and 1017 rise by about +0.08 and 1002 by about +0.05. The step holds until 11:40. Same pattern on 22.10 (11:10), 27.10 (11:15) and 02.11 (11:20); not on 03.11.
- **Images**: fixed-frame thumbnails of 8012 at 10:01, 11:00 and 12:01, and on 31.10 at 12:00: nothing in front, nothing that changes visibly at this size. Whole images of 02.11 and 03.11 at 11:01: sun both days. No image between 11:00 and 12:01.
- **Confidence**: medium. The fact is sure (the reference changes brightness and shifts, the base follows). The cause is only probable, since 03.11, also sunny, does not show it.
### 2026-11-01T11:15Z → 2026-11-01T11:25Z
- **Map phenomenon**: Atmospheric turbulence, or a partial passage like that of 11:00; I could not decide
- **Family**: Visibility, from the image-quality signature (sharpness down, `dominance` up, flux whole: the image shakes, conventions of `lire-les-donnees`). Choosing Obstruction would have required an image of the passage.
- **Measured**: `quality.py --table sharpness` at 5 min: median sharpness at 0.81 at 11:20, 0.79 at 11:30 and 0.73 at 11:45, against 0.89 to 0.93 between these measurements; 8014, 8017 and 1021 to 1024 are almost not touched, as at 11:00. At 11:20, the pose makes a one-measurement round trip (tilt at 0.0149°, against 0.0104° and 0.0087° around; common `x` at −0.52 against −0.38 and −0.34). `ranges.py --rolling 3 --instants 1`: 6 sensors between 3 and 6 noises at 11:20 (`corr_x` of 7013H, 1024, 102H, 1010, 92H; `corr_y` of 1017), only one at 11:25; none at 11:30.
- **Images**: none at 11:20 (hourly images only).
- **Confidence**: low for the cause, strong for the fact (a one-measurement jump).
## 2. What I left without a label
- 11:30 and 11:35: the image is even less sharp (11:30) and the reference 8013 at times loses part of its light, but `ranges.py` shows no sensor off level on `corr_x` or `corr_y`. No effect on the curves.
- 10:30 to 10:50: `quality.py --outside` flags 8010 (and, weakly, 101B, 1005, 1007), because of the `dominance` of 8010, high all morning (3 to 4.6 times its calm reference) without change; its `corr_x` stays within 3 noises. It is a usual state at this hour, not an event of the batch.
- The `error_value` of the reference 8017 above its median from 10:30 to 10:40, and the `corr_y` of 8017 and 1024 a little low at the start of the batch: it is the slow evolution of the morning cycle, already under the thermal label.
- The thermal movement of the camera is labeled (D-117), even where the compensation absorbs it; so nothing thermal is left aside.
## 3. The documentation
- **What served me**: the LENZSPITZE variant in "A pass-by hides the targets…" (flux whole, `dominance` up, a second pattern); the turbulence sheet (flux whole, image less sharp); "The sun… behind a spot" (the background of a reference in the sun makes a small common jump); the LENZSPITZE line of section 4; the rule of `families.md` that puts a reference moving away with a sharp edge in Visibility; the bound rules at the edge of the batch.
- **What I missed**: on a station whose images are taken only hourly, nothing says how to separate an intermittent turbulence from a partial passage when the same sensors lose their sharpness. The quality signature is the same, and the map gives no spatial criterion (e.g. the sensors to the right of the field intact). The map also does not know a reference that *gains* light at once: 8012 brightens at the same hour on several sunny days, and not all.
- **Apparent contradiction**: a passage is given as "one measurement, sometimes two" that comes back at the next. Here, the passage falls at the turning point of the thermal cycle (the `corr_y` of several targets makes a step that stays). The `corr_y` over the whole day was needed to see that this step is the return from the morning dip, and not a trace of the train.
- No indication from the expert in conversation, apart from the hint of the request (a train around 11:00), confirmed by the image.
## Difficulties
- No clear category for a one-measurement jump whose cause hesitates between turbulence and partial passage, without an image; I put it in Visibility by the signature, with the reservation written.
- The crosses of `thumbnails.py` are slightly offset from the spots on LENZSPITZE (reconstructed frame, `displacement-columns.md`): to keep in mind so as not to read a spot "out of its place".
## Sources read
- `.claude/skills/lire-les-donnees/SKILL.md`
- `labeling/label-map.md`: summary, sections 2 to 4, sheets "A pass-by hides the targets for the length of one measurement", "A target left under the threshold latches onto a headlight, a reflection or another spot", "A vehicle or machine that stays in front of the targets", "Atmospheric turbulence", "The sun in the field or behind a spot", "The camera moves through a thermal effect", "The camera drifts away when the temperature of its housing drops, without a daily cycle", "A step of the camera that the compensation does not erase"
- `labeling/families.md`; `labeling/labeling-rules.md`
- `labeling/knowledge/measurement-chain.md`, `displacement-columns.md`, `gaps.md`, `quality-metrics.md`
- `labeling/auto/LENZSPITZE.jsonl` (neighboring labels and habits of the station)
- Data: `data/LENZSPITZE/` through `inventory.py`, `overview.py` (batch, day, `--rigid`), `metrics.py`, `quality.py` (`--common`, `--table`, `--figure`, `--labels --outside`), `ranges.py`, `summary.py`, `plot.py`, `thumbnails.py`, the raw CSV of 01.11 (instants), my script `scripts/capteur_jours.py`; images of 01.11 at 10:01, 11:00 and 12:01, of 02.11 and 03.11 at 11:01
Binary file not shown.

Before

Width:  |  Height:  |  Size: 158 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 321 KiB

@@ -1,4 +0,0 @@
{"start": "2026-11-01T10:30Z", "end": "2026-11-01T11:40Z", "family": "Thermal displacement", "description": "Part of the day's thermal cycle, in the sun (11:00 image: full sun on the track, sharp shadows); the cycle started early in the morning and continues after the end of the batch. During the batch, the housing temperature rises markedly and the camera moves: the tilt, which has risen since the morning, drops at once at 11:05, the pan drops at once at 11:15, then the rotation rises little by little; the scale stays roughly stable. All the targets and the references move together in the image. The corr_y of many targets, dipped since the morning, comes back toward its night level from 11:00, in one step then slowly (1020, 1019, 1018, 1016, 1008, 1024, 1023, 1021): the compensation leaves a remainder of these camera movements. Deduced cause: the pose follows the housing temperature, as on every sunny day on this station, and the curves come back with it in the afternoon. Normal, nothing to do. The 11:00 train, the reference 8012 that lights up at 11:15 and the image that shakes at 11:20 are labelled separately."}
{"start": "2026-11-01T10:55Z", "end": "2026-11-01T11:05Z", "family": "Obstruction", "description": "A train passes on the track to the left of the instrumented track, seen as translucent, motion blur, on the 11:00 image: it passes in front of 8003, 8010, the reference 8004 and the bottom of the distant references 8011, 8012 and 8013. At the measurements of 11:00 and 11:01, the image of almost all sensors, references included, is less sharp and ambiguous (a second pattern shows up in it, especially on 8003, 8010, 8004, 1016, 1014, 1012 and 101B), but their light stays whole, except that of 1006, partly hidden: the train was in front of them for only part of the capture. The error_value of almost all sensors rises markedly, without any rejection; the curves make a small jump for the time of these two measurements: the corr_y of 1022 especially, of 202H, 1009, 1013, 1001, 1005, 102H and 1020, the corr_x of 101B, 1012, 1013, 1008 and of the reference 8019; everything comes back at 11:05. The passage falls at the start of the return of the thermal cycle (separate label), which at the same time makes the corr_y of several targets rise. Cause seen on the 11:00 image. The tracks are not moving: ignore these points."}
{"start": "2026-11-01T11:05Z", "end": "2026-11-01T11:40Z", "family": "Visibility", "description": "At 11:15, the spot of the distant reference 8012 lights up at once: its light increases markedly, its image becomes sharper and less ambiguous, and its position in the image shifts relative to the other references and the targets; it stays so until the end of the batch. No reference is set aside, but the recalibration error of the camera rises and stays high: the basis changes. After a first deviation at 11:10, then a step at 11:15, the corr_x of 8012 rises strongly and that of the targets shifts according to their place in the image: 8003, 1011, 1001, 1003, 1006, 1021, 201B, 7013H and the reference 8004 drop, 8010, 1017 and 1002 rise; 8003, 8010 and 1017 also shift a little in the image. The camera turns at the same moment (thermal label). The same sudden light of 8012, with the same shift, comes back around the same hour on 22.10, 27.10 and 02.11, sunny days, but not on 03.11, which was sunny too. Cause not established, probably the sun reaching the reference 8012 or its background at that hour; no image between 11:00 and 12:01. This is not a movement of the tracks: do not trust the shift of the corr_x of the targets over the range."}
{"start": "2026-11-01T11:15Z", "end": "2026-11-01T11:25Z", "family": "Visibility", "description": "At 11:20, the image of almost all sensors is less sharp and more ambiguous, without loss of light; the references 8014 and 8017, on the right of the field, and the near targets 1021 to 1024 are little affected. The camera pose makes a one-measurement round trip (tilt) and the corr_x of 7013H, 1024, 102H, 1010 and of the reference 92H, the corr_y of 1017, make a small jump, then everything comes back at 11:25. The same disturbance, weaker, comes back at 11:30 with no effect on the curves. Image shaking in the sun, deduced from the image quality (light whole, sharpness dropping): heat shimmer above the sunlit track is probable, but it touches the same sensors as the 11:00 train and a partial passage is not excluded; no image at that moment. This is not a movement: ignore this point."}
Binary file not shown.

Before

Width:  |  Height:  |  Size: 407 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 177 KiB

@@ -1,61 +0,0 @@
# LENZSPITZE: a train at 11:00, then a reference that brightens
You selected 01.11 from 10:30 to 11:40 UTC with this question: "Around 11:00, the track targets move a little; a train?" Yes, a train passes at 11:00, seen on the image. But what the curves keep afterwards, from 11:15, has another cause: the distant reference 8012 changes brightness and the base of the compensation moves with it. All this happens in the middle of the camera's thermal cycle, on a sunny day.
## What the data show
There is no rejection, no missing measurement, and the 9 references stay retained over the whole range; the camera is re-registered at each measurement (`inventory.py --start … --end …`).
**1. 10:55 → 11:05: a train passes (confirmed).**
- *What is seen.* On the image of 11:00, a train passes on the track to the left of the instrumented track, in transparency, with motion blur. It covers 8003, 8010 and the reference 8004, and passes at the bottom of the distant references 8011 to 8013.
- *Duration.* The images, taken once an hour, do not give it: the train is only on that of 11:00 (absent at 10:01 and at 12:01), moving during the capture; it is the measurements that bound it: it touches those of 11:00 and 11:01, not those of 10:55 or 11:05: it stayed in front of the targets at least one minute, and less than ten (a long train, or two passages).
- *Image quality.* At the measurements of 11:00 and 11:01, the image of almost all sensors, references included, becomes less sharp, and a second pattern shows in it. The light remains whole, except on 1006, partly hidden: the train was in front of the targets only part of the capture (`quality.py --table sharpness|dominance|flux --step 1min`).
- *What does not move.* The references 8014 and 8017, to the right of the field, and the near targets 1021 to 1024 are barely touched.
- *Error and curves.* The `error_value` of 35 sensors out of 41 rises, without any rejection (`overview.py --min 1`). The curves make a small jump for the time of these two measurements: the `corr_y` of 1022 mostly, then 202H, 1009, 1013 and 1001, and the `corr_x` of 101B, 1012, 1013, 1008 and of the reference 8019. Everything comes back at 11:05 (`ranges.py --rolling 3 --instants`, `summary.py --step 1min`).
**2. 10:30 → 11:40 (and beyond): the camera moves with the heat (strongly supported).**
- *Temperature.* The housing temperature rises clearly from 11:00 and peaks in the early afternoon (`metrics.py`).
- *Camera pose.* The tilt drops at once at 11:05, the pan at 11:15, then the rotation rises. All targets and references move together in the image (`overview.py`, `--rigid`).
- *Curves.* The `corr_y` of many targets (1020, 1019, 1018, 1016, 1008…), dipped since the morning warming, rises from 11:00 and regains its night level around 13:00 (`summary.py --step 30min`, from 03:00 to 16:00). Part of the "movement" seen around 11:00 is therefore this return, not the train: it starts at the same instant, but it lasts.
- *Same days.* The same cycle returns every sunny day on this station.
**3. 11:05 → 11:40: the reference 8012 brightens and slightly distorts the base (consistent with the sun on this reference).**
- *The reference.* At 11:15, the spot of 8012 suddenly becomes brighter (about one and a half times) and sharper. Its position in the image shifts by about a third of a pixel against the other references, which stay under a tenth of a pixel (`capteur_jours.py`, `overview.py --rigid 2026-11-01T10:30Z`).
- *The base.* The registration error (`sigma0`) doubles and stays high.
- *The curves.* The `corr_x` of the targets makes a step depending on their place in the image: 8003, 1011, 1001, 1003, 1006, 1021, 201B, 7013H drop; 8010, 1017, 1002 rise. The step holds until the end of the range.
- *The other days.* The same brightening of 8012 returns around the same hour on 22.10, 27.10 and 02.11, sunny days, but not on 03.11, also sunny.
**4. 11:20: the image shakes for one measurement (possible: turbulence; a partial passage is not excluded).**
- *Image quality.* The image loses sharpness on the same sensors as at 11:00, without loss of light.
- *Pose and curves.* The pose makes a round trip and a few targets make a small jump (7013H, 1024, 102H, 1010, 1017, the reference 92H), which comes back at 11:25. The same nuisance returns weaker at 11:30, without effect on the curves.
![Corrected curves from 10:00 to 12:30; grey zone: the passage of the train](courbes-corr.png)
In the grey zone, 1022 and 202H make a one-measurement jump in `corr_y` and come back. Just after, 1020 and 1018 rise and stay (the thermal return). At 11:15, 8012 (yellow) leaps in `corr_x`, and 8003, 1011 drop while 8010 and 1017 rise: this is the step of the base.
![The camera image at 11:00](image-1100.jpg)
A train is seen to the left of the instrumented track, in transparency and blurred, under full sun.
![Track-target zone at 10:01, 11:00 and 12:01](vignettes-zone-voie.png)
At 11:00, the train passes in front of 8003, 8004 and the group 1002, 1004, 1006. It is neither at 10:01 nor at 12:01.
![Image quality from 10:00 to 12:30 (relative to calm days; colored ranges: the labels)](qualite-commun.png)
The sharpness drops and the `dominance` leaps at 11:00, while the flux stays whole: the train does not hide the spots, it blurs them. The weaker dips of 11:20, 11:30 and 11:45 are also visible.
![Camera pose and raw positions over the range](pose-et-positions.png)
The pan drops and `sigma0` rises at 11:15; on the `x` panel, the reference 8012 (purple) stands out from the others.
## What remains not established
- **Why 8012 brightens.** The sun on the reference or on its background is likely: the hour repeats from one sunny day to the next. But 03.11, sunny, does not show it, and no image exists between 11:00 and 12:01.
- **What the step of 11:15 distorts.** The step of the `corr_x` mixes two things: the base distorted by 8012, and a small own shift of 8003, 8010 and 1017 in the image. I did not separate them (the compensation would have to be redone without 8012).
- **The cause of 11:20.** Turbulence above the sunlit track or partial passage of a train: the image quality does not tell them apart, and no image is taken at that instant.
- **The camera steps.** The steps of the tilt (11:05) and of the pan (11:15) are more abrupt than an ordinary cycle. I read them as the thermal cycle, because they return in the afternoon; a link with the train is not demonstrated.
## Labels
No existing label covered this range. I propose four:
- `2026-11-01T10:30Z → 11:40Z` **Thermal displacement**: the camera moves with the warming of the housing, in the sun; the `corr_y` of many targets comes back from its morning dip.
- `2026-11-01T10:55Z → 11:05Z` **Obstruction**: a train seen on the image of 11:00. Images blurred without loss of light (except 1006), small jumps of one or two measurements, return at 11:05.
- `2026-11-01T11:05Z → 11:40Z` **Visibility**: the reference 8012 brightens and shifts, and the base of the compensation is distorted by it. The `corr_x` of the targets makes a step; probable cause the sun, not established.
- `2026-11-01T11:15Z → 11:25Z` **Visibility**: at 11:20, the image shakes for one measurement, with small jumps that come back. Probable turbulence, partial passage not excluded.
@@ -1,6 +0,0 @@
{
"station": "LENZSPITZE",
"start": "2026-11-01T10:30Z",
"end": "2026-11-01T11:40Z",
"hints": "Vers 11:00, les cibles de voie bougent un peu ; un train ?"
}
@@ -1,19 +0,0 @@
"""Un capteur, plusieurs jours, mêmes heures : flux, dominance, x, corr_x, au pas natif.
Usage : python3 capteur_jours.py CAPTEUR HH:MM HH:MM JOUR [JOUR ...] (JOUR = AAAAMMJJ)
Lit data/LENZSPITZE/displacement/ ; n'écrit rien.
"""
import csv
import sys
capteur, h1, h2, *jours = sys.argv[1:]
for jour in jours:
chemin = f"data/LENZSPITZE/displacement/LENZSPITZE_advanced_{jour}.csv"
print(f"--- {capteur} {jour}")
for r in csv.DictReader(open(chemin)):
t = r["time_utc"][11:16]
if r["sensor_name"] == capteur and h1 <= t <= h2:
def f(c, n=3):
return f"{float(r[c]):.{n}f}" if r[c] else "-"
print(t, "flux", f("flux", 0), "dom", f("dominance"), "x", f("x", 2),
"cx", f("corr_x"), "ev", f("error_value"))
Binary file not shown.

Before

Width:  |  Height:  |  Size: 318 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 821 KiB

@@ -1,134 +0,0 @@
# Automatic labels: NORDEND, 2026-11-04T21:12Z → 2026-11-07T14:12Z (UTC)
Analysis requested from the Data page (hint: "the detectors do not see the disturbances in corr-x and corr-y"). No label of `labeling/auto/NORDEND.jsonl` touches the range: the last ones end on 01/11 at 00:00, and the batch from 01 to 04/11 is not labeled. Nothing is therefore extended, nor contradicted. Working figures in `work/`, those of the report at the root of the folder.
## 1. My labels
### 2026-11-04T21:15Z → 2026-11-05T06:30Z
- **Map phenomenon**: The camera moves through a thermal effect
- **Family**: Thermal displacement. The pose follows the housing temperature and comes back with it; it is always labeled.
- **Measured**: night of 04 to 05/11, housing temperature from 22.8 to 19.2 °C; pan rising (0.028 → 0.039°), tilt dropping (−0.141 → −0.152°), scale dropping, without rest until the turning point of 05/11 at 06:30 (pan 0.0389, then 0.0379 at 06:45, tilt −0.1508 then −0.1493). `corr_x` of the right-edge targets a little above the day (`scripts/cycle_jour.py`). The cycle started on 04/11, before the batch, with no neighboring label: it starts at the batch edge.
- **Images**: none useful (night, web, nothing that touches the pose).
- **Confidence**: medium. The start of this cycle is outside the batch.
### 2026-11-04T21:15Z → 2026-11-07T14:00Z
- **Map phenomenon**: A target that degrades, and sometimes recovers (CTR01, continuation of NORDEND 16.10 19:00)
- **Family**: Target issue, like the reviewed label NORDEND-195 of the October batch (rule: same target, same family).
- **Measured**: CTR01 rejected at 124 measurements out of 260; median error from 0.42 to 0.38 per day (threshold 0.4), as since 16/10 (`scripts/par_jour.py`, from 20/09 to 12/11). Rejections almost all night from 04 to 05/11, then every morning until early afternoon (05/11 07:00-14:45, 06/11 06:45-14:45, 07/11 06:30-14:00). Outside the compensation (`pattern_type` = `ref`): its rejections do nothing to the other targets.
- **Images**: thumbnails of 04/11 23:00, 05/11 10:00 and 22:00, 06/11 11:00, 07/11 04:00 and 12:00 against PR17. Small spot at night, almost invisible by day when PR17 stays sharp.
- **Confidence**: medium. Old and constant regime; the cause remains not established.
### 2026-11-05T00:45Z → 2026-11-05T01:30Z
- **Map phenomenon**: A spider web in front of the lens (a thread on a reference), with A reference discarded: the base changes
- **Family**: Obstruction. A seen thread hides a reference (boundary "A hindered reference"; same family as NORDEND 18.08 22:15, thread on CTR03).
- **Measured**: CTR03 at 0.295 and 0.356 (00:30, 00:45), rejected at 01:00 (0.404) and 01:15 (0.553), 0.280 at 01:30. At both rejections, 0 reference retained (`inlier` empty for CTR02 and CTR04), reprojection of CTR02 and CTR04 at zero, `sigma0` empty; 17 targets out of 18 more than 4 noises in `corr_y` (median −0.073 and −0.079), 11 in `corr_x` (+0.015). PR17 and PR18 at 28 and 26 noises. Nothing at 00:45 nor at 01:30 (`scripts/pics_ctr03.py`). CTR02 and CTR04 stay under their threshold.
- **Images**: zone around the references at 01:00: a big blurred bright point of the thread on the spot of CTR03. Whole image of 05/11 23:00 for the web.
- **Confidence**: strong.
### 2026-11-05T03:00Z → 2026-11-05T09:45Z
- **Map phenomenon**: A spider web in front of the lens (a thread on a spot)
- **Family**: Obstruction. A single target, a thread seen in front of it at night.
- **Measured**: error of PR09 at 0.29 in median over the morning (0.10 to 0.12 on the other days), rejected at 03:15, 05:45, 06:15, 07:15, 07:45, 08:15, 08:30, 08:45 and 09:30; PR10 and PR11 at their level. The `corr` of PR09 does not move outside the rejections of CTR03 (`ranges.py --rolling 8`). Bounds: the measurement before the first rejection, the measurement after the last.
- **Images**: `--fixed` thumbnails of PR09 and PR10: at 06:00, the halo of the thread covers the background of PR09; at 08:00 (day), nothing visible; the night of 04/11 23:00, nothing.
- **Confidence**: medium. The thread is seen at night; by day, the cause is deduced.
### 2026-11-05T03:15Z → 2026-11-05T03:45Z
- **Map phenomenon**: A spider web in front of the lens; A reference discarded: the base changes
- **Family**: Obstruction. Same reference, same signature, wandering thread seen at 01:00 and 04:00.
- **Measured**: CTR03 rejected at 03:30 only (0.534); 0 reference retained; 17 targets out of 18 jump in `corr_y` (median −0.081), return at 03:45. PR09 is rejected at 03:15, the start bound: it is the other event (PR09 label), which overlaps this one.
- **Images**: no image at 03:30 (one per hour); at 04:00, the thread is between CTR03 and CTR02.
- **Confidence**: medium (cause deduced).
### 2026-11-05T06:30Z → 2026-11-06T07:00Z
- **Map phenomenon**: The camera moves through a thermal effect
- **Family**: Thermal displacement.
- **Measured**: morning turning point at 06:30 (pan and tilt leave their night level at 06:45). Housing temperature from 19.2 to 29.7 °C (max at 14:00). Pan lowest 0.0099° around 14:15 (−44 noises of the night), tilt highest around 11:30-12:00, rotation lowest at 14:00, scale highest at 14:45 (`days.py --night 3-5 --bounds`). Slow return in the evening and at night, without rest until the turning point of 06/11 at 07:00. Day minus night, in noises: `corr_x` PR18 −7, PR16 −5, PR13, PR14, PR15, PR17 −3; `corr_y` PR18 and PR17 +6, PR16 +5. Jolts: 09:45 (8 targets a little low in `corr_y`, in the middle of the fast tilt movement) and 12:00. At 12:00, the raw median `y` of all sensors, references included, takes +0.43 px for one measurement, the rotation and the tilt jump: 12 targets more than 4 noises in `corr_y` (median −0.020), return at 12:15. Placed in the cycle label (rule of the pose agitation).
- **Images**: none needed. The image of 06/11 12:00 shows an overcast day, without obstacle in front of the group.
- **Confidence**: strong for the cycle, medium for the choice of keeping the jolt of 12:00 in the cycle.
### 2026-11-05T14:45Z → 2026-11-05T15:15Z
- **Map phenomenon**: A spider web in front of the lens (a thread on a spot)
- **Family**: Obstruction. Same target, same cause as in the morning; separated from the morning by more than five hours without rejection.
- **Measured**: PR09 rejected at 15:00 (0.410), neighbors at their level; PR09 still right at the threshold in the afternoon (high error range 14:15-19:30).
- **Images**: thumbnail of 15:00: nothing visible (day).
- **Confidence**: low to medium (cause deduced).
### 2026-11-05T23:00Z → 2026-11-06T10:00Z
- **Map phenomenon**: A spider web in front of the lens (placed on a reference, it makes it move away); A reference discarded: the base changes
- **Family**: Obstruction. Thread seen on CTR03.
- **Measured**: the error of CTR03 goes from 0.12 (19:30) to 0.21 (19:45) then 0.32 (20:00) and stays between 0.25 and 0.40. Rejected at 13 measurements, from 23:15 to 09:45. At each, 0 reference retained, reprojection of CTR02 and CTR04 at zero, 15 to 18 targets out of 18 more than 4 noises in `corr_y` (median −0.065 to −0.080). Between the rejections, nothing clear: at most 3 targets off level. Start at the measurement before the first rejection, and not at 19:45. From 19:45 to 23:00, CTR03 accepted shifts the targets only under their noise: against the targets, CTR03 moves away by about 0.01 px (`overview.py --rigid`), and `cib_corr_y` rises by 2 noises at most.
- **Images**: thumbnails and zone at 21:00, 23:00, 06/11 06:00: a wide bright blurred band of the thread on CTR03; CTR02 and CTR04 sharp. Whole image of 23:00 (`fig2`).
- **Confidence**: strong.
### 2026-11-06T07:00Z → 2026-11-07T06:30Z
- **Map phenomenon**: The camera moves through a thermal effect
- **Family**: Thermal displacement.
- **Measured**: turning point at 07:00 (pan 0.0383 then 0.0374 at 07:30; the poses of 06:00, 06:45, 07:15 on two references are left out of the reading). Housing up to 27.2 °C. Pan lowest around 13:45 (−39 noises), tilt highest around 13:00, scale highest at 13:00. Slow return, incomplete on the morning of 07/11: pan lower by 11 noises, tilt higher by 16 noises than the night before; the night is warmer (housing around 22 °C against 19 °C). Day minus night: `corr_x` PR18 −10, PR16 −8, PR15 −6, PR13 −5, PR17 and PR14 −4, PR04 +7; `corr_y` PR18 +8, PR16, PR17, PR15, PR08 +7.
- **Images**: 06/11 12:00, overcast day, nothing in front of the group.
- **Confidence**: strong.
### 2026-11-06T09:15Z → 2026-11-06T09:45Z
- **Map phenomenon**: A spider web in front of the lens (a thread on a spot)
- **Family**: Obstruction. A single target; the thread seen on it the night before.
- **Measured**: error of PR10 from 0.24 to 0.26 at 09:00-09:15, rejected at 09:30 (0.419), 0.095 at 09:45. Nothing on its neighbors. This label falls under that of CTR03 (rejected at 09:45, the end bound): two events, two sensors, overlap intended ("what falls under a long label is labeled apart").
- **Images**: thumbnail of 05/11 22:00: halo of the thread on PR10. On 06/11 by day, nothing.
- **Confidence**: low to medium.
### 2026-11-06T21:15Z → 2026-11-07T11:00Z
- **Map phenomenon**: A spider web in front of the lens; A reference discarded: the base changes
- **Family**: Obstruction.
- **Measured**: CTR03 between 0.18 and 0.30 all of 06/11 by day (without rejection or effect on the curves), then rejected at 22 measurements from 21:30 to 10:45, tighter in the morning (09:30 to 10:15 in a row). At each, 0 reference retained, 15 to 17 targets in `corr_y` (median −0.058 to −0.075). After 10:45, CTR03 stays under the threshold (0.28 to 0.40) until 12:15, without effect on the curves. It falls back to 0.108 at 12:30.
- **Images**: thumbnail of 07/11 04:00: the bright band on CTR03; 07/11 10:00 (day): nothing visible; 07/11 20:00 (after the batch): the thread moved to the right of CTR03.
- **Confidence**: strong.
### 2026-11-07T06:30Z → 2026-11-07T14:00Z
- **Map phenomenon**: The camera moves through a thermal effect
- **Family**: Thermal displacement.
- **Measured**: turning point at 06:30 (pan 0.0296 then 0.0286, tilt −0.1412 then −0.1382). Housing from 22 to 26.3 °C, less hot than the days before. Pan lowest around 13:00, tilt highest around 10:15, rotation lowest around 09:45, scale rising. Not back at the end of the batch. Day minus night: `corr_x` PR18 −7, PR13 −6, PR16 −4; `corr_y` PR06 +17, PR16 +8, PR15 +6. The reference night is low for PR06 (−0.02 the night of 06 to 07/11).
- **Images**: 07/11 14:00, the web against the light on the group (`fig5`); nothing else.
- **Confidence**: medium. The cycle is cut by the end of the batch, and the reference night is mixed with the rejections of CTR03.
### 2026-11-07T11:15Z → 2026-11-07T11:45Z
- **Map phenomenon**: A spider web in front of the lens (a thread on a spot), cause deduced
- **Family**: Obstruction.
- **Measured**: error of PR04 rising since 09:30 (0.15 to 0.19), rejected at 11:30 (0.733), 0.129 at 11:45; neighbors at their level; `corr` of PR04 without a range off level apart from this point.
- **Images**: thumbnails of PR04 at 11:00 and 12:00: whole spot, background a little brighter, nothing clear. Image of 14:00: the web crosses the group.
- **Confidence**: low (cause deduced from a thread seen near the group that day, not on PR04).
### 2026-11-07T13:45Z → 2026-11-07T14:00Z
- **Map phenomenon**: A spider web in front of the lens (a thread on a spot)
- **Family**: Obstruction.
- **Measured**: PR07 rejected at 14:00 (0.401), last measurement of the batch; `end` at the batch edge, the next measurement (14:15) being outside the selection. Over 6 measurements (12:00-14:00), its error is higher than usual.
- **Images**: thumbnail and image of 14:00: the background of PR07 lit by a bright band of the thread that descends through the group (`fig5`).
- **Confidence**: medium.
## 2. What I left without a label
- **CTR02 right at the threshold on 05/11, from 01:45 to 16:15** (error between 0.08 and 0.38, without rejection): the thread of the web passes over it (zone at 04:00). No target moves at these instants: at most 1 to 3 targets more than 4 noises from their local level (`pics_ctr03.py`). An error that rises without effect on `corr` is not labeled.
- **CTR03 accepted but veiled** (05/11 19:45-23:00, 06/11 by day, 07/11 11:00-12:15): barely shifted against the targets, without effect on the `corr` beyond their noise.
- **Error rises of the targets without rejection or jump**: PR01 and PR02 on 05/11 (night and morning), PR11 during the nights of CTR03, PR12 and PR17 on 07/11 at the end of the morning (PR17 up to 0.29 for a threshold of 0.4). `ranges.py --rolling 8` finds no own deviation of `corr` for them.
- **The web itself**, present the whole batch: apart from the named sensors, it does not veil the scene. The error of the other targets stays at its level; it is not a Visibility.
- **The dirty glass (NORDEND 25.10 05:45, closed on 01/11)**: no step nor change of `corr` level linked in this batch; not taken up.
- **The thermal movement** is labeled every day, as D-117 requires.
## 3. The documentation
- **What served**: the sheet "A spider web in front of the lens", especially "placed on a reference, it makes it move away: all targets on a step" (NORDEND 18.08 22:15), and the boundary "A hindered reference" of `families.md`, which give the family. The sheet "A reference dropped: the base changes" for the mechanism (reprojection at zero). The expert's rule "three references are needed". The sheet of the NORDEND thermal cycle, without rest at night, and the morning turning point rule.
- **What I missed**:
- A word on the **comb**: a reference right at the threshold for hours, rejected one measurement in two or three. The sheet says "step" (a reference hidden and rejected in a row). Here, they are one-measurement peaks, repeated all night, to be cut into one episode per night rather than one event per rejection. I followed the rule "one label per parking or per episode". The threshold of the "pause of several measurements" that cuts an episode is not quantified: I cut at the day without rejection (06/11) and kept together two-hour pauses at night.
- How to bound an episode whose nuisance (CTR03 right at the threshold) starts hours before the first touched point. I bounded on the curves (first rejection), as the bound rule asks.
- **What seemed contradicted**: nothing.
## Difficulties
- The instruction asks to read the whole batch, but the batch from 01 to 04/11 has no label. The thermal cycle and the regime of CTR01 therefore start at the batch edge, with no neighboring label to extend.
- `days.py --bounds`: the reference night is mixed with the peaks of CTR03 (the measurements of the targets there remain accepted). The night level of the `corr` is distorted by it: `corr_y` 07/11 "96 measurements beyond the threshold". I reread the cycles with my own script, which leaves out the instants where CTR03 is rejected (`scripts/cycle_jour.py`).
## Sources read
- `.claude/skills/lire-les-donnees/SKILL.md`
- `labeling/label-map.md`: summary, sections 2 to 4, sheets B (l. 119-290: passage, headlight, vehicle, animal, web, plant, obstacle on the spot, dirty glass, rain, visibility, part of the scene, reference that loses its brightness, turbulence), C, D and E (l. 293-434)
- `labeling/families.md`, `labeling/labeling-rules.md` (in full)
- `labeling/knowledge/`: `measurement-chain.md`, `displacement-columns.md`, `gaps.md`, `thermal-cycle.md`, `camera-images.md`, `quality-metrics.md`, `expert-annotated-cases.md`
- `labeling/auto/NORDEND.jsonl` (the labels of the October batch that touch CTR01, CTR03, PR09, the web and the thermal cycle)
- Data: `data/NORDEND/` through `inventory.py`, `overview.py` (with `--rigid`), `metrics.py`, `days.py --bounds`, `ranges.py`, `summary.py`, `plot.py`, `thumbnails.py`, the whole images of 05/11 18:00 and 23:00, 06/11 12:00, 07/11 14:00; the raw CSVs through my scripts `scripts/pics_ctr03.py`, `scripts/par_jour.py`, `scripts/cycle_jour.py`, `scripts/bornes.py`
Binary file not shown.

Before

Width:  |  Height:  |  Size: 187 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 854 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 1.6 MiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 686 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 618 KiB

@@ -1,14 +0,0 @@
{"start": "2026-11-04T21:15Z", "end": "2026-11-05T06:30Z", "family": "Thermal displacement", "description": "End of the camera thermal cycle of 04/11, from the start of the batch to the turn of the morning: at night, as the housing cools, the camera slowly comes back, the pan rises markedly, the tilt drops by the same amount, the scale drops; it does not rest before the morning. The targets on the right edge (PR18, PR16, PR13 to PR17) rise back a little in corr_x with it; the rest is absorbed by the compensation. The jumps at 01:00, 01:15 and 03:30 come from the reference CTR03 (separate labels). Deduced cause: the pose follows the housing temperature, and the quarry's rock face does not move with temperature. The structure is not moving, nothing to do."}
{"start": "2026-11-04T21:15Z", "end": "2026-11-07T14:00Z", "family": "Target issue", "description": "CTR01 (local reference, read as a target) stays right at its threshold throughout the batch, as since 16/10: rejected for a large part of the time, almost all night from 04 to 05/11 then mostly by day, every morning until early afternoon; its curve is missing a large part of the time and its kept points are very noisy. It does not enter the compensation: its rejections do not make the other targets jump. On the thumbnails, its spot is small at night and almost invisible by day, while that of PR17 stays sharp (05/11 10:00, 06/11 11:00, 07/11 12:00). Cause not established: state of the prism, or of the window in front of it, to be checked. Ignore its curve; act on the target."}
{"start": "2026-11-05T00:45Z", "end": "2026-11-05T01:30Z", "family": "Obstruction", "description": "A thread of the cobweb stretched in front of the lens, lit by the infrared and beaded, passes over the reference CTR03 (01:00 thumbnail: a large, blurry, bright dot covers its spot, CTR02 and CTR04 stay sharp). CTR03 is rejected at 01:00 and 01:15: the compensation then runs only on CTR02 and CTR04 (their reprojection falls to zero, no more sigma0), the basis changes and almost all targets jump together, markedly downward in corr_y, a little upward in corr_x, PR17, PR18 and PR11 the most; the raw positions do not move. Everything comes back at 01:30. Cause seen on the thumbnail. Very bad sign for these two measurements: do not trust the curves over the range; the structure is not moving."}
{"start": "2026-11-05T03:00Z", "end": "2026-11-05T09:45Z", "family": "Obstruction", "description": "PR09 right at its threshold all morning, rejected several times from 03:15 to 09:30: its curve has gaps. At night, the halo of a thread of the cobweb, lit by the infrared, passes over its spot (06:00 thumbnail), then over its neighbours at the bottom left; by day the thread is no longer visible but PR09 stays right at the threshold. Its neighbours PR10 and PR11 stay measured. Cause seen at night on the thumbnail, deduced by day. Ignore the curve of PR09 over the range; the structure is not moving."}
{"start": "2026-11-05T03:15Z", "end": "2026-11-05T03:45Z", "family": "Obstruction", "description": "The reference CTR03 is rejected for one measurement, at 03:30: the compensation runs on CTR02 and CTR04 alone (reprojection at zero), and almost all targets jump together, markedly downward in corr_y, PR17, PR18 and PR13 the most, then come back at the next measurement. Deduced cause: the thread of the cobweb that wanders that night between CTR03 and CTR02 (01:00 and 04:00 thumbnails); there is no image at 03:30. Ignore the points at 03:30; the structure is not moving."}
{"start": "2026-11-05T06:30Z", "end": "2026-11-06T07:00Z", "family": "Thermal displacement", "description": "Camera thermal cycle: in the morning, when the housing heats up (markedly, highest around 14:00), the camera turns: the pan drops strongly until early afternoon, the tilt rises markedly, the rotation drops a little, the scale increases; then it comes back in the evening and all night, as the housing cools, with no rest until the start of the next morning. In the middle of the movement, the pose fluctuates from one measurement to the next: at 09:45, then at 12:00 when the whole scene, references included, makes a one-measurement jolt, many targets drop a little in corr_y and come back at once. The compensation absorbs it poorly on the right edge: the corr_x of PR18 dips markedly around noon, PR16, PR17, PR13 to PR15 a little; the corr_y of PR18, PR17 and PR16 rises a little; the curves come back in the evening. The jumps of the night come from CTR03 (separate label). Deduced cause: the pose follows the housing temperature, and the rock face does not move with temperature. The structure is not moving, nothing to do."}
{"start": "2026-11-05T14:45Z", "end": "2026-11-05T15:15Z", "family": "Obstruction", "description": "PR09, still right at its threshold in the afternoon, is rejected for one measurement at 15:00: one point is missing, its neighbours stay measured. Deduced cause: the thread of the cobweb seen on its spot in the night (label of the morning), invisible by day. Ignore this point; the structure is not moving."}
{"start": "2026-11-05T23:00Z", "end": "2026-11-06T10:00Z", "family": "Obstruction", "description": "A thread of the cobweb stretched in front of the lens, lit by the infrared, settled on the reference CTR03 in the early evening (21:00, 23:00 and 06:00 thumbnails: a wide, bright, blurry band over its spot, CTR02 and CTR04 sharp next to it); its error_value then goes very close to the threshold and stays there. CTR03 is rejected at many measurements, from 23:15 to 09:45: at each one, the compensation runs only on CTR02 and CTR04 (reprojection at zero, no more sigma0), the pose flips and almost all targets jump together, markedly downward in corr_y and a little upward in corr_x, PR17, PR18, PR11 and PR13 the most, then come back at the next measurement: the curves look like a comb of spikes. The raw positions do not move. The disturbance continues by day without rejection (label of the following night separate). Cause seen on the images. Very bad sign for these measurements: do not trust the points of the range; the structure is not moving. The thread has to be removed from the window."}
{"start": "2026-11-06T07:00Z", "end": "2026-11-07T06:30Z", "family": "Thermal displacement", "description": "Camera thermal cycle: in the morning, when the housing heats up (markedly, highest in early afternoon), the camera turns: the pan drops strongly, the tilt rises markedly, the rotation drops a little, the scale increases; then it slowly comes back in the afternoon and at night, as the housing cools less than the day before, without coming all the way back: the next morning, the pan stays lower and the tilt higher than at the start. The compensation absorbs it poorly on the right edge: the corr_x of PR18 dips markedly in the afternoon, PR16, PR15 and PR13 a little less, PR17 and PR14 a little; PR04 rises a little; the corr_y of PR18, PR16, PR17, PR15 and PR08 rises a little. The spikes of the morning and of the night come from CTR03 (separate labels). Deduced cause: the pose follows the housing temperature, and the rock face does not move with temperature. The structure is not moving, nothing to do."}
{"start": "2026-11-06T09:15Z", "end": "2026-11-06T09:45Z", "family": "Obstruction", "description": "PR10 rises toward its threshold in late morning and is rejected for one measurement at 09:30: one point is missing, its neighbours stay measured, and the next point has already come back. Deduced cause: the thread of the cobweb, seen on its spot in the night (05/11 22:00 thumbnail), invisible by day. Ignore this point; the structure is not moving."}
{"start": "2026-11-06T21:15Z", "end": "2026-11-07T11:00Z", "family": "Obstruction", "description": "Second night of the cobweb thread on the reference CTR03 (the first is labelled separately, on 05/11 at 23:00): the bright, blurry band, lit by the infrared, still covers its spot (04:00 thumbnail); CTR03, which stayed near the threshold all day, is rejected at many measurements from 21:30 to 10:45, more and more often toward the morning. At each rejection, the compensation runs only on CTR02 and CTR04 (reprojection at zero), the pose flips and almost all targets jump together, markedly downward in corr_y and a little upward in corr_x, PR17, PR18, PR11 and PR13 the most, then come back at the next measurement. The raw positions do not move. After 11:00, CTR03 stays under the threshold and no longer disturbs the curves; at 12:30 its error_value falls back to its level, and in the evening the thread has moved beside its spot (07/11 20:00 image). Cause seen on the images. Very bad sign for these measurements: do not trust the points of the range; the structure is not moving. The thread has to be removed from the window."}
{"start": "2026-11-07T06:30Z", "end": "2026-11-07T14:00Z", "family": "Thermal displacement", "description": "Camera thermal cycle, until the end of the batch: in the morning, when the housing warms up (less than the days before), the camera turns: the pan drops markedly, the tilt rises markedly, the rotation drops, the scale increases; at the end of the batch it has not yet come back. The compensation absorbs it poorly: the corr_x of PR18 and PR13 dips a little, PR16 a little less; the corr_y of PR06 rises markedly, that of PR16 and PR15 a little. The spikes until 10:45 come from CTR03 (separate label). Deduced cause: the pose follows the housing temperature, and the rock face does not move with temperature. The structure is not moving, nothing to do."}
{"start": "2026-11-07T11:15Z", "end": "2026-11-07T11:45Z", "family": "Obstruction", "description": "PR04, whose error_value has been rising since late morning, is rejected for one measurement at 11:30, markedly above the threshold: one point is missing, its neighbours stay measured and the next point has come back. Deduced cause: the thread of the cobweb, which passes over the group of targets that day (bright band on the 14:00 image); nothing is visible on its 11:00 and 12:00 thumbnails. Ignore this point; the structure is not moving."}
{"start": "2026-11-07T13:45Z", "end": "2026-11-07T14:00Z", "family": "Obstruction", "description": "PR07 is rejected at 14:00, the last measurement of the batch, just above its threshold: one point is missing. On the 14:00 image, a bright, blurry band (a lit thread of the cobweb) goes down through the group of targets very close to it and lights the background of its thumbnail. Cause seen on the image. Ignore this point; the structure is not moving."}
@@ -1,84 +0,0 @@
# NORDEND: a spider web on the reference CTR03 makes all the curves jump
Selection: from 04/11 at 21:12 to 07/11 at 14:12 (UTC). Hint from the user: the detectors do not see the disturbances of `corr_x` and `corr_y`. These disturbances are a comb of peaks, mostly at night: at each peak, almost all targets jump together, for one measurement, then come back.
## What the data show
### 1. The peaks: the reference CTR03 rejected, the compensation falls to two references
- **When**: 38 measurements, all isolated instants or pairs: on 05/11 at 01:00, 01:15 and 03:30; the night of 05 to 06/11, from 23:15 to 09:45 (13 measurements); the night of 06 to 07/11, from 21:30 to 10:45 (22 measurements).
- **What moves**: at each of these instants, and only there, the reference CTR03 goes above its error threshold (0.40 to 0.58 for a threshold of 0.4) and the station discards it. The compensation then runs only on CTR02 and CTR04: their reprojection error falls to zero and `sigma0` is missing. The base changes and **15 to 18 targets out of 18** jump together: `corr_y` downward (median of the targets about −0.07 to −0.08, i.e. 20 to 30 times the noise for PR17 and PR18), `corr_x` a little upward (about +0.01 to +0.02). The most touched targets are PR17, PR18, PR11 and PR13. Everything comes back at the next measurement.
- Commands: `python3 tools/scripts/inventory.py data/NORDEND --start … --end …` (rejections of CTR03 and "0 kept; outside: CTR02 (empty), CTR03 (r), CTR04 (empty)" at the 38 instants); `scripts/pics_ctr03.py` (for each instant, the error of CTR03 and the number of targets more than 4 noises from their local level: 15 to 18 in `corr_y` at the 38 rejections, never more than 3 elsewhere, except at two instants of 05/11 read below).
- **What does not move**: the raw positions. Against the common movement of the targets, no target jumps at the rejections (`overview.py --rigid 2026-11-05T18:00Z`, residual of the targets of the order of a hundredth of a pixel). CTR02 and CTR04 stay under their threshold. No other target is rejected at these instants.
- **Level of proof of the mechanism**: **confirmed** (reference rejected, reprojection at zero, common jumps at the same instants, nothing in the raw positions).
![CTR03 against three targets: error_value at the top, corr_y at the bottom; grey zones: the labels of CTR03](fig1_ctr03_corr_y.png)
At the top, the error of CTR03 (blue) rises at once on 05/11 around 19:45 and lives right at the threshold until 07/11 around 12:30; at the bottom, each time it passes the threshold, PR17, PR18 and PR11 dive together for one measurement. Between two rejections, their curves are at their level.
![Reprojection and residuals of the references against the targets (overview.py --rigid)](fig4_rigide_reprojection.png)
Bottom panel: the reprojection of CTR02 and CTR04 falls to zero at each rejection of CTR03 (green crosses at the top); the compensation then runs on two references. Top panels: CTR03 moves slightly away from the targets during its two nights; the targets stay a block.
### 2. The cause: a thread of spider web in front of the lens, on CTR03
- A web stretched very close to the lens crosses the image in wide blurred bands, lit at night by the infrared, beaded with drops. It is there the whole batch. The wind moves it from one night to the next.
- On 05/11 at 01:00, a big bright blurred point of the thread sits on CTR03: this is the rejected measurement. At 04:00, it has moved toward CTR02; the error of CTR02 then rises until the morning, without rejection or effect on the curves.
- On 05/11 from 19:00-21:00, a band of the thread comes onto CTR03 and stays there two nights (thumbnails of 21:00, 23:00, 06/11 06:00, 07/11 04:00). The error of CTR03 goes from about 0.1 to 0.3-0.4 and stays right at the threshold, also by day (thread not lit, invisible). On 07/11 at 12:30, the error drops back; in the evening, the thread has moved beside the spot (image of 07/11 at 20:00).
- The same web bothers a few targets, without touching the other curves: PR09 on 05/11 in the morning (halo on its spot at 06:00, rejected several times from 03:15 to 09:30, then at 15:00), PR10 on 06/11 at 09:30, PR04 on 07/11 at 11:30, PR07 on 07/11 at 14:00 (bright band very close, image of 14:00). Only their rejected points are missing.
- Commands: `thumbnails.py --sensors CTR03,CTR02,CTR04` and `--zone 1550,700,700,600`; whole images by `ls data/NORDEND/images/ | grep …` then reading.
- **Level of proof**: **confirmed** for the thread on CTR03 the nights of 05 and 06/11 and at 01:00 on 05/11 (seen on the images of those hours); **strongly supported** for 03:30 on 05/11 and for the daytime rejections (no image at the instant, or thread invisible by day; same reference, same signature, thread seen the night before and after).
![Whole image of 05/11 at 23:00: the web in front of the lens](fig2_image_20261105T2300Z.jpg)
The bright blurred bands are the threads of the web, very close to the lens. The group of targets is at the bottom right; CTR03 is the highest point to the left of the group, under the band that descends toward the bottom right.
![The zone of the references CTR03 and CTR02, from 04/11 22:00 to 07/11 20:00](fig3_fil_sur_CTR03.png)
The thread moves: on CTR03 on 05/11 at 01:00, between the two at 04:00, on CTR03 at 21:00 (and all the following night), invisible by day (06/11 09:00, 07/11 10:00), then to the right of CTR03 on 07/11 at 20:00, when CTR03 has come back to its level.
![Whole image of 07/11 at 14:00: the web seen by day](fig5_image_20261107T1400Z.jpg)
By day, the web is seen against the light: a bright band descends through the group of targets, with a big bright point. It is at this hour that PR07 is rejected.
### 3. What also moves, unrelated to the web: the thermal cycle of the camera
- Every day, when the housing heats (housing temperature highest around 14:00 on 05/11, early afternoon on 06/11), the camera turns: the pan drops strongly, the tilt rises, the scale increases. It comes back slowly in the evening and all night, without rest. Morning turning points: 05/11 06:30, 06/11 07:00, 07/11 06:30.
- The compensation almost absorbs it, except on the right edge of the group: the `corr_x` of PR18 dips clearly in the middle of the day (7 to 10 times the noise), that of PR16 and PR13 to PR17 a little less.
- On 05/11 at 12:00, the camera makes a one-measurement jolt: the whole scene, references included, moves in the image. Many targets then drop a little in `corr_y` then come back. I put it in the cycle of the day.
- Commands: `metrics.py` (housing temperature, pose, `cib_corr_x`), `days.py --night 3-5 --bounds`, `scripts/cycle_jour.py`.
- **Level of proof**: **strongly supported** (the pose follows the housing temperature; the rock face does not move with the temperature, according to the station context).
### 4. What is missing: CTR01
CTR01 (local reference, read as a target, outside the compensation) is rejected a large part of the batch (124 measurements out of 260), as since 16/10. Its spot is small at night and almost invisible by day, when that of PR17 stays sharp. Its rejections do not make the other targets jump. Cause not established (**insufficient evidence**): state of the prism, or of the glass in front of it.
## Why the detectors see it badly (hypothesis)
Each peak lasts one measurement, and the next step is back. They also do not go through a rejected measurement of a target: the targets jump under their threshold, and only the reference is rejected. A detector that looks for ranges of rejections of the targets, or for deviations that last, does not see them. The signature to look for is a **reference rejected with the reprojection of the other two at zero**, and almost all targets jumping at the same instant. **possible**, not verified on the code of the detectors (outside what I am allowed to read).
## What remains not established
- Why CTR03 passes the threshold at these precise instants rather than at the neighboring ones: the thread covers it all night, its error oscillates around the threshold; the exact instant probably depends on the movement of the thread (wind), invisible at the rate of one image per hour.
- Between two rejections, CTR03 accepted but veiled is slightly shifted against the targets (of the order of a hundredth of a pixel): its effect on the pose is under the noise of the curves; I did not label it.
- The state of CTR01 (prism or glass); the nature of the rejections of PR04 at 11:30 on 07/11 (nothing on the thumbnail, cause deduced).
- After the batch: PR07 is rejected almost without stopping from 08 to 10/11, and the measurements are missing from 11/11 00:15 to 12/11 10:30. Outside the selection, not read.
## Labels
- 04/11 21:15 → 05/11 06:30, Thermal displacement: end of the cycle of 04/11, the camera comes back at night.
- 04/11 21:15 → 07/11 14:00, Target issue: CTR01 right at the threshold the whole batch, curve with holes.
- 05/11 00:45 → 01:30, Obstruction: thread of the web on CTR03 (seen at 01:00), CTR03 discarded, all targets jump.
- 05/11 03:00 → 09:45, Obstruction: thread of the web on PR09 (seen at 06:00), rejected points.
- 05/11 03:15 → 03:45, Obstruction: CTR03 discarded for one measurement (03:30), all targets jump; thread deduced.
- 05/11 06:30 → 06/11 07:00, Thermal displacement: camera cycle, with the jolt of 12:00.
- 05/11 14:45 → 15:15, Obstruction: PR09 rejected for one measurement, thread deduced.
- 05/11 23:00 → 06/11 10:00, Obstruction: thread on CTR03, first night, comb of peaks on all targets.
- 06/11 07:00 → 07/11 06:30, Thermal displacement: camera cycle.
- 06/11 09:15 → 09:45, Obstruction: PR10 rejected for one measurement, thread deduced.
- 06/11 21:15 → 07/11 11:00, Obstruction: thread on CTR03, second night, comb of peaks on all targets.
- 07/11 06:30 → 14:00, Thermal displacement: camera cycle, until the end of the batch.
- 07/11 11:15 → 11:45, Obstruction: PR04 rejected for one measurement, thread deduced.
- 07/11 13:45 → 14:00, Obstruction: PR07 rejected at 14:00, thread seen on the image.
For the client: the structure does not move. The peaks of the two nights are not valid measurements: the station had only two references left. The web in front of the lens must be removed (glass to clean); the problem will return as long as it stays.
@@ -1,6 +0,0 @@
{
"station": "NORDEND",
"start": "2026-11-04T21:12Z",
"end": "2026-11-07T14:12Z",
"hints": "les détecteurs ne voient pas les perturbations en corr-x et corr-y"
}
@@ -1,22 +0,0 @@
"""Vérifie les bornes : instants de mesure, debut < fin, et à chaque borne les
capteurs rejetés (CTR01 exclu, rejeté d'habitude) et CTR03.
Usage : python3 bornes.py data/NORDEND etiquettes.jsonl
"""
import csv, glob, json, sys
dossier, fichier = sys.argv[1], sys.argv[2]
rows = {}
for f in sorted(glob.glob(f'{dossier}/displacement/*_advanced_202611*.csv')):
for r in csv.DictReader(open(f)):
t = r['time_utc'][:16].replace(' ', 'T') + 'Z'
t = t[:14] + f"{int(t[14:16]) - int(t[14:16]) % 15:02d}" + 'Z'
rows.setdefault(t, {})[r['sensor_name']] = r
for n, l in enumerate(open(fichier), 1):
d = json.loads(l)
ok = d['debut'] < d['fin'] and d['debut'] in rows and d['fin'] in rows
info = []
for b in ('debut', 'fin'):
rj = [c for c, r in rows.get(d[b], {}).items()
if c != 'CTR01' and float(r['error_value']) > float(r['error_threshold'])]
if rj: info.append(f"{b} rejetés : {','.join(rj)}")
print(f"{n:2} {'OK ' if ok else 'ERR'} {d['debut']} → {d['fin']} {d['categorie']:<20} {'; '.join(info)}")
@@ -1,30 +0,0 @@
"""Par jour et par cible : corr de la journée (médiane 10:00-16:00) moins la nuit
d'avant (médiane 02:00-05:00), en bruits, sans les instants où CTR03 est rejetée.
Usage : python3 cycle_jour.py data/NORDEND AAAA-MM-JJ[,AAAA-MM-JJ...]
"""
import csv, glob, sys, statistics as st
dossier, jours = sys.argv[1], sys.argv[2].split(',')
rows = {}
for f in sorted(glob.glob(f'{dossier}/displacement/*_advanced_*.csv')):
for r in csv.DictReader(open(f)):
if r['time_utc'][:10] in jours:
rows.setdefault(r['time_utc'][:16], {})[r['sensor_name']] = r
rej = lambda r: float(r['error_value']) > float(r['error_threshold'])
bad = {t for t, d in rows.items() if 'CTR03' in d and rej(d['CTR03'])}
cibles = sorted({n for d in rows.values() for n in d if n.startswith('PR')})
for j in jours:
print(f'== {j} : jour 10-16 moins nuit 02-05 (bruits), triés')
for c in ('corr_x', 'corr_y'):
out = []
for n in cibles:
def serie(h0, h1):
return [float(d[n][c]) for t, d in sorted(rows.items()) if t[:10] == j and h0 <= t[11:13] < h1
and t not in bad and n in d and not rej(d[n]) and d[n][c]]
nuit, jour = serie('02', '05'), serie('10', '16')
tout = serie('00', '24')
if len(nuit) < 3 or len(jour) < 3: continue
b = st.median(abs(a - b) for a, b in zip(tout, tout[1:])) or 1e-9
out.append(((st.median(jour) - st.median(nuit)) / b, n))
out.sort()
print(f' {c}: ' + ' '.join(f'{n}:{v:+.0f}' for v, n in out))
@@ -1,17 +0,0 @@
"""Par jour : error_value médiane et nombre de rejets de capteurs choisis.
Usage : python3 par_jour.py data/NORDEND AAAAMMJJ AAAAMMJJ CAPT1,CAPT2
"""
import csv, glob, sys, statistics as st
dossier, d0, d1, capteurs = sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4].split(',')
print('jour ' + ''.join(f'{c:>22}' for c in capteurs))
for f in sorted(glob.glob(f'{dossier}/displacement/*_advanced_*.csv')):
jour = f[-12:-4]
if not d0 <= jour <= d1: continue
ev = {c: [] for c in capteurs}; rj = {c: 0 for c in capteurs}
for r in csv.DictReader(open(f)):
c = r['sensor_name']
if c in ev:
v = float(r['error_value']); ev[c].append(v)
rj[c] += v > float(r['error_threshold'])
print(jour + ' ' + ''.join(f"{(st.median(ev[c]) if ev[c] else 0):>8.3f} {rj[c]:>4}/{len(ev[c]):<4} " for c in capteurs))
@@ -1,64 +0,0 @@
"""Pics des corr des cibles face aux rejets de CTR03, instant par instant.
Niveau local de chaque cible : médiane de ses corr dans ±2 h, sans les instants
où CTR03 est rejetée. Imprime, par instant : error_value de CTR03 (r si rejetée),
CTR02, CTR04, sigma0, médiane des résidus corr_x / corr_y des cibles acceptées,
nombre de cibles à plus de K bruits (bruit = médiane des |écarts| d'un pas).
Usage : python3 pics_ctr03.py data/NORDEND DEBUT FIN [K]
"""
import csv, glob, sys, statistics as st
from datetime import datetime, timedelta
dossier, debut, fin = sys.argv[1], sys.argv[2], sys.argv[3]
K = float(sys.argv[4]) if len(sys.argv) > 4 else 4
t0 = datetime.fromisoformat(debut.replace('Z', '+00:00'))
t1 = datetime.fromisoformat(fin.replace('Z', '+00:00'))
REFS = ('CTR02', 'CTR03', 'CTR04')
rows = {} # t -> name -> dict
for f in sorted(glob.glob(f'{dossier}/displacement/*_advanced_*.csv')):
jour = f[-12:-4]
if not (t0 - timedelta(days=1)).strftime('%Y%m%d') <= jour <= (t1 + timedelta(days=1)).strftime('%Y%m%d'):
continue
for r in csv.DictReader(open(f)):
t = datetime.fromisoformat(r['time_utc']).replace(second=0, microsecond=0)
t = t.replace(minute=t.minute - t.minute % 15)
if t0 - timedelta(hours=3) <= t <= t1 + timedelta(hours=3):
rows.setdefault(t, {})[r['sensor_name']] = r
T = sorted(rows)
def rej(r): return float(r['error_value']) > float(r['error_threshold'])
bad = {t for t in T if 'CTR03' in rows[t] and rej(rows[t]['CTR03'])}
cibles = sorted({n for t in T for n in rows[t] if n not in REFS and n != 'CTR01'})
def val(t, n, c):
r = rows[t].get(n)
if r is None or rej(r) or r[c] == '': return None
return float(r[c])
res = {} # (t, n, c) -> residu en bruits
for c in ('corr_x', 'corr_y'):
for n in cibles:
serie = [(t, val(t, n, c)) for t in T]
ok = [v for t, v in serie if v is not None and t not in bad]
bruit = st.median(abs(a - b) for a, b in zip(ok, ok[1:])) or 1e-9
for i, (t, v) in enumerate(serie):
if v is None: continue
voisins = [w for (u, w) in serie[max(0, i - 8):i + 9] if w is not None and u not in bad and u != t]
if voisins:
res[(t, n, c)] = ((v - st.median(voisins)), bruit)
print('instant CTR03 CTR02 CTR04 sigma0 med_dcx med_dcy n>K(cx) n>K(cy) top')
for t in T:
if not t0 <= t <= t1: continue
ev = {n: rows[t][n] for n in REFS if n in rows[t]}
s = lambda n: ('r' if rej(ev[n]) else '') + f"{float(ev[n]['error_value']):.3f}" if n in ev else '-'
sig = next((rows[t][n]['sigma0'] for n in rows[t] if rows[t][n]['sigma0']), '')
line = []
for c in ('corr_x', 'corr_y'):
d = [res[(t, n, c)] for n in cibles if (t, n, c) in res]
line.append((st.median(x for x, _ in d) if d else 0, sum(abs(x) > K * b for x, b in d), d))
top = sorted(((abs(res[(t, n, 'corr_y')][0]) / res[(t, n, 'corr_y')][1], n) for n in cibles if (t, n, 'corr_y') in res), reverse=True)[:3]
print(f"{t:%Y-%m-%dT%H:%MZ} {s('CTR03'):>6} {s('CTR02'):>6} {s('CTR04'):>6} {sig[:7]:>7} "
f"{line[0][0]:+.4f} {line[1][0]:+.4f} {line[0][1]:>4} {line[1][1]:>4} "
+ ' '.join(f'{n}:{v:.0f}' for v, n in top))
@@ -1,52 +0,0 @@
# Labels: NORDEND, 2026-11-05T06:00Z → 2026-11-05T12:00Z (UTC)
The range is already covered by four labels of `labeling/auto/NORDEND.jsonl`, which I keep and do not redo:
- NORDEND-229 (Thermal displacement, 04/11 21:15 → 05/11 06:30): the end of the night cycle, up to the morning turning point;
- NORDEND-234 (Thermal displacement, 05/11 06:30 → 06/11 07:00): the thermal cycle of the day;
- NORDEND-230 (Target issue, CTR01, 04/11 21:15 → 07/11 14:00): CTR01 rejected at 23 of the 25 measurements of the range;
- NORDEND-232 (Obstruction, PR09, 05/11 03:00 → 09:45): the rejections of PR09 (06:15, 07:15, 07:45, 08:15 to 08:45, 09:30).
I add a single label, the camera jolt at 12:00.
## 1. My labels
### 2026-11-05T11:45Z → 2026-11-05T12:15Z
- **Map phenomenon**: A camera step that the compensation does not erase.
- **Precedents**: NORDEND-80 (01/09 15:30, Displacement, reviewed): same shape, a one-measurement step of the pose, several targets that jump together and come back. I took up its family, its one-measurement bounds on either side and its conclusion. SNCF01-42 (15/10 05:15): a step that stays, also classed Displacement. MICHIGAN-2: a one-measurement deviation, classed Thermal because the housing temperature makes a brief drop there, which is not the case here (temperature in regular rise, without a jolt over the hour).
- **Family**: Displacement. It is not the thermal agitation, because the pose does not make a larger step in the direction of the morning drift: it jumps (rotation, tilt), then comes back right away at the next measurement. `families.md`, "what is not labeled": a camera step that leaves something on the `corr` is labeled, and concludes "the camera moves". It is not Visibility (no rise of the `error_value` specific to the instant, sharp image) nor Obstruction (no reference lost, nothing in front). NORDEND-234 already describes the jolt in its text; the rule "what falls under a long label is labeled apart" gives it its own label.
- **Measured**:
- camera pose (`overview.py`, 15 min table): rotation −0.0096° at 11:45, −0.0054° at 12:00, −0.0106° at 12:15; tilt −0.1209°, −0.1161°, −0.1229°; sigma0 from 0.00209 to 0.00181 then 0.00211;
- raw positions: median of `y` at +0.77, +1.21, +0.74 px; median of `x` at −0.15, −0.25, −0.12 px; the three references with the targets (CTR04 in `y` up to +1.28);
- `--rigid` (base 06:00): fitted rotation from −0.0092° to −0.0031° then −0.0089°; residual of CTR04 at +0.032 / +0.029 px, above its morning level; `cib_max` at 0.060 px against about 0.04 before and after;
- `ranges.py --rolling 4 --instants 3`: 8 sensors off level at 12:00, and only at that instant. In `corr_y`: PR01 −7, PR07 −6, PR10 −6, PR02 −5, PR04 −5 noises; in `corr_x`: CTR04 +6. The `corr` of the references also move (CTR02 in `corr_y` −6, CTR03 +6);
- `summary.py`, `corr_y`: PR01 0.162 → 0.123 → 0.161; PR10 0.188 → 0.148 → 0.184; PR04 0.139 → 0.103 → 0.128; PR07 0.102 → 0.072 → 0.102; PR02 0.170 → 0.135 → 0.158; everything comes back at 12:15;
- what does not move: no reference lost (3 retained at the 25 instants), no rejection outside CTR01 (usual regime). The `error_value` of PR01, CTR02 and PR09 is high at 12:00 (0.23 to 0.25), but these three sensors go like this from a low level to a high level all morning (10:15 to 10:45, 11:15), without a pose jolt or effect on the `corr`: it is not the cause. The housing temperature rises regularly (about 25.5 °C at 11:00, 27.8 °C at 12:00, hourly median).
- **Images**: the whole image of 12:00 shows a sunny day, the left of the scene in the sun, the targets in the shade, nothing in front of them, only a blurred bright streak in the sky, far from the targets. Thumbnails of PR09, PR10, PR12, PR01, CTR03 and PR17 at 06:00, 09:00, 10:00 and 12:00: sharp spots by day, the same at 10:00 and at 12:00.
- **Confidence**: medium. The shape is clear and passes the threshold of 5 noises on 8 sensors, but the deviation is small (a few hundredths of a px). The cause is not seen. A second jolt of the same shape follows at 12:30 (rotation at −0.0034°, tilt at −0.1309°), outside the range and not read.
## 2. What I left without a label
- **09:45, one-measurement dip in `corr_y` of PR09, PR10, PR12 (−5 to −7 noises), weaker on PR03, PR05, PR13, PR02, PR01; return at 10:00.** The pose makes no jolt there: it is the moment when the morning thermal drift goes fastest (tilt +0.0055° in one step, against +0.002 to +0.003 around), and it continues in the same direction without coming back. The `error_value` drops instead of rising. According to the rule "the agitation from one measurement to the next is written in the thermal label", it stays in NORDEND-234, whose description already says it ("the pose shakes from one measurement to the next: at 09:45"). It is the most fragile boundary of my work: a reviewer could make it a second one-measurement Displacement.
- **The drift of the right edge** (`corr_x` of PR18, PR16, PR15, PR13, PR14 dropping, `corr_y` of PR17, PR18, PR16 rising) **and the slow rise of the `corr_x` of PR04 and the `corr_y` of PR08** follow the pose from 06:30 to noon: it is the thermal cycle, NORDEND-234. Its description does not name PR04 or PR08 for this day; they are not separate events.
- **CTR01**: rejected almost everywhere, covered by NORDEND-230.
- **PR09**: the rejections until 09:30 are covered by NORDEND-232. Its end at 09:45 is perhaps one step too early: PR09 is still low in `corr_y` at 10:00 (0.126 against about 0.15 to 0.17), i.e. about 3 to 4 of its noises, and back at 10:15. Its dip of 09:45 is shared with PR10 and PR12 (above). I do not change it: the deviation of 10:00 stays at the edge of its noise.
- **The back-and-forth of the `error_value` of CTR02 (reference), PR01 and PR09** between a low level and a high level, under the threshold except the rejections of PR09: it does not touch the `corr` at those instants (10:15, 10:30, 10:45, 11:15). It is not labeled.
## 3. The documentation
- **What served**: the sheet "A step of the camera that the compensation does not erase" and the precedent NORDEND-80, for the family and the conclusion; the boundary of `families.md` on the camera step that leaves a remainder in the `corr`; rules 1 ("what falls under a long label is labeled apart") and 5 (agitation in the thermal label), to separate 12:00 from 09:45; the NORDEND context of section 4 (rock face, camera cycle every day, web).
- **What I missed**: a clear rule between "agitation of the pose, in the thermal label" and "one-measurement camera step, labeled apart", when the jolt falls in the middle of a cycle. I decided by the shape of the pose: a step that comes back at the next measurement is a jolt, a larger step in the direction of the drift is agitation. This criterion is mine, not the map's.
- **What seemed contradicted**: NORDEND-234 puts the jolt of 12:00 in the thermal cycle, without its own label. For me, it requires one (above). The rest of NORDEND-234 is right on the range: start at 06:30, at the turning point of the tilt and the pan, verified on the pose table.
- **Detectors**: not read (DETECTORS: no).
## Difficulties
- `labeling/precedents.py` failed at first: Ollama did not answer. The container then started by itself ("health: starting"), and the search worked on rerun.
## Sources read
- `instructions/analyst.md`, `tools/GUIDE.md`
- `labeling/label-map.md`: summary, sections 2 to 4, sheets "The camera moves through a thermal effect", "The camera drifts away when the temperature of its housing drops, without a daily cycle", "A step of the camera that the compensation does not erase"
- `labeling/families.md`, `labeling/labeling-rules.md` (in full)
- `labeling/knowledge/measurement-chain.md`, `displacement-columns.md`, `gaps.md`
- `labeling/auto/NORDEND.jsonl` (labels from 04 to 07/11 and Target issue of CTR01)
- precedents: two `precedents.py` queries (NORDEND-80, SNCF01-42, MICHIGAN-2, MICHIGAN-8, LENZSPITZE-45 and other thermal cycles)
- tools: `inventory.py`, `overview.py` (with and without `--rigid 2026-11-05T06:00Z`), `ranges.py --rolling 4 --instants 3`, `summary.py` (`corr_y`, `error_value`), `metrics.py`, `plot.py`, `thumbnails.py`, image of 12:00
Binary file not shown.

Before

Width:  |  Height:  |  Size: 135 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 376 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 417 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 509 KiB

@@ -1 +0,0 @@
{"start": "2026-11-05T11:45Z", "end": "2026-11-05T12:15Z", "family": "Displacement", "description": "For the duration of one measurement, at 12:00, in the middle of the camera thermal cycle (labelled separately), the pose makes a jolt: the rotation and the tilt jump then come back at the next measurement, and the raw positions of all sensors, references included, jump together in the image then come back, not quite by the same amount (CTR04 a little apart from the targets). The compensation does not quite erase it: the corr_y of PR01, PR10, PR02, PR04 and PR07 drops markedly, PR11, PR09 and PR17 a little, and the corr_x of PR01 and PR02 rises a little; everything comes back at 12:15. No reference is lost (all three stay in the compensation) and no target is rejected apart from CTR01, rejected as usual; the error_value of the reference CTR02 and of PR01, which comes and goes all morning with no effect on the curves, rises clearly at 12:00 with the jolt, under the threshold, and comes back at 12:15; that of PR09 barely moves. On the 12:00 image, sunny day, the targets in shadow, nothing in front of them. Cause not established: a jolt of the camera or of its mount, deduced from the pose. This is not a movement of the structure; points to ignore, the camera is to be watched if it repeats."}
@@ -1,47 +0,0 @@
# NORDEND: a morning of thermal cycle, and a camera jolt at noon
The selection runs from 05/11 06:00 to 05/11 12:00 (UTC), without hints. The question: what do the corrected curves show over this morning, and what is not yet labeled?
## What the data show
- **06:30 → noon (and until the next day): the thermal cycle of the camera.** Already labeled (NORDEND-234).
- The housing heats from about 19 °C to 28 °C. The camera turns: the pan drops, the tilt rises, the scale increases.
- The whole scene, references included, slides together in the image. The compensation almost absorbs it: a drift of a few hundredths of a px remains on the targets of the right edge (PR18, PR16, PR15, PR13 in `corr_x`) and on PR04 and PR08.
- Cause: **strongly supported**. The pose follows the housing temperature, and the rock face does not move with the temperature.
- Commands: `overview.py data/NORDEND --start 2026-11-05T04:00Z --end 2026-11-05T14:00Z --min 1`; `metrics.py` on the same window.
- **12:00, one measurement: a camera jolt that the compensation does not erase.** New label, Displacement.
- The rotation and the tilt jump then come back at 12:15. The raw positions of all sensors, references included, jump together then come back.
- The `corr_y` of PR01, PR10, PR02, PR04 and PR07 drops clearly, by 5 to 7 times their noise. The `corr_x` of PR01 and PR02 rises a little. Everything comes back at 12:15.
- What does not move: no reference lost, no rejection outside CTR01 (as usual), nothing in front of the targets on the image of 12:00.
- Cause: a jolt of the camera or its mount, **consistent with** the pose; not seen, therefore not established.
- Commands: `ranges.py data/NORDEND --start 2026-11-05T04:00Z --end 2026-11-05T14:00Z --columns corr_x,corr_y --rolling 4 --instants 3` (8 sensors off level at 12:00 only); `overview.py ... --rigid 2026-11-05T06:00Z`.
- **The missing points**: all already labeled.
- CTR01 is rejected over almost the whole range: Target issue, NORDEND-230.
- PR09 is rejected at times until 09:30: thread of the spider web seen at night, NORDEND-232.
- Command: `inventory.py data/NORDEND --start 2026-11-05T06:00Z --end 2026-11-05T12:00Z`.
- **09:45, one measurement**: PR09, PR10 and PR12 drop a little in `corr_y`, then come back.
- It is the moment when the thermal drift goes fastest, without a pose jolt.
- I leave it in the thermal cycle, whose label already cites it.
![Corrected curves around noon](fig-acoup-12h.png)
The `corr_y` of PR01, PR02, PR04, PR07 and PR10 dips at 12:00 (grey zone) and comes back at 12:15. The small dip of 09:45 is also visible, weaker.
![Camera pose and raw positions](fig-pose-commun.png)
At the top, the pose drifts all morning with the heat (thermal cycle). At 12:00, the rotation, the tilt and the raw position of all sensors make a one-measurement peak, then come back.
![Camera image at 12:00](image-1200.jpg)
Sunny day. The targets are in the shade, at the bottom right, and nothing is in front of them.
![Residuals of corr_x](fig-residus-corr_x.png)
The slow drift of the right edge (PR18, PR16, PR15, PR13) over the morning: it is the rest of the thermal cycle.
## What remains not established
- The cause of the jolt of 12:00: a brief movement of the camera or its mount, without an image of the instant that shows it. A second jolt of the same shape follows at 12:30, outside the selection and not read.
- The end of NORDEND-232 (PR09) is perhaps one step too early: PR09 is still a little low at 10:00.
- The back-and-forth of the `error_value` of CTR02, PR01 and PR09 all morning, without effect on the curves, remains unexplained.
## Labels
- 05/11 11:45 → 12:15, Displacement: one-measurement jolt of the camera at 12:00, several targets drop in `corr_y` and come back; not a movement of the structure, points to ignore, camera to monitor if it repeats.
- Kept as they are: NORDEND-229, NORDEND-234 (Thermal displacement), NORDEND-230 (Target issue, CTR01), NORDEND-232 (Obstruction, PR09).
@@ -1,6 +0,0 @@
{
"station": "NORDEND",
"start": "2026-11-05T06:00Z",
"end": "2026-11-05T12:00Z",
"hints": ""
}
@@ -1,29 +0,0 @@
# Tools: NORDEND_2026-11-05T0600_2026-11-05T1200
**Tools modified**: none
**data.py**: unchanged
## What the analyst used
- `inventory.py` on the range: references kept at each instant, rejections of CTR01 and PR09.
- `overview.py --min 1` (widened window from 04:00 to 14:00), then `--rigid 2026-11-05T06:00Z --follow … --residuals no`: the pose and the table by step dated the jolt of 12:00 (rotation, tilt, raw positions); the residual PNGs showed the drift of the right edge.
- `ranges.py --columns corr_x,corr_y --rolling 4 --instants 3`: under the thermal drift, it returned only two instants (09:45 with 3 sensors, 12:00 with 8), with the state of the references. It is the command that grounds the label.
- `summary.py --step 15min --level window --mark` (`corr_y`, `error_value`), `metrics.py`, `plot.py --zone`, `thumbnails.py --radius 40`, the whole image of 12:00.
- The folder has no `scripts/`: the analyst wrote no script. Its "Difficulties" mention only Ollama (`labeling/precedents.py`), which does not depend on this repository. Its "Documentation" section talks about the labeling rules, not about reading.
- What was cluttering, slightly: a `| tail -22` on `summary.py` and a `| head -20` on `metrics.py`, to cut an output. `--start`/`--end` and the pipe to `grep` described in the guide already do it.
## What I decided
- All the tools used: left as they are. They showed what the analyst had to see: the one-measurement jolt under the drift (`ranges.py --rolling --instants`), the pose and the references (`overview.py`, `--rigid`). Nothing was missing for it.
- `summary.py`, `metrics.py`: left as they are. The `head`/`tail` were the analyst's choice, not a lack of the tool. The Conventions of the guide already say how to shorten an output.
- No script to bring in: `scripts/` does not exist.
## What changed
none
## Trial
No tool changed, so there was nothing new to try. I nevertheless reran, from `tools/scripts/`, the tools the report cites: the analysis ran just before commit 67c8862, which moved the tools into `scripts/`. The figures are stored in `work/`: common, multiples-error_value, residuals-corr_x/corr_y/error_value, rigid, metrics.
## Report commands rerun
5, with the `tools/scripts/` paths (`overview.py --min 1`, `metrics.py`, `ranges.py --rolling 4 --instants 3`, `overview.py --rigid 2026-11-05T06:00Z`, `inventory.py`): no failure. `ranges.py` returns the same two instants (09:45: 3 sensors; 12:00: 8 sensors) as those cited by the analyst.
## Difficulties
- The log cites the tools at their old path (`tools/inventaire.py`, `tools/ensemble.py` …). The report names them without a path. I reran with `tools/scripts/`, like the current guide.
@@ -0,0 +1,78 @@
# Labels: WYOMING, 2026-11-06T08:05Z → 2026-11-16T06:27Z (UTC)
Measurements of the range: 2026-11-06T08:15Z → 2026-11-16T06:15Z, 15 min, 951 instants. 16 sensors: 3 references (RD, target_12, target_13), 13 targets. No image quality metrics (export columns peak to snr empty). One image per hour, on the hour. Temperature: only `temperature_deg_c` is filled; I call it the housing temperature (`labeling-rules.md` §5). No earlier labels (`labeling/auto/WYOMING.jsonl` does not exist).
**Revision after the user's remark on the weather.** I added the MeteoSwiss hourly weather at Genève-Cointrin (station GVE, 4.9 km from "Genève", the place given for WYOMING in `~/aimsight-data/extraction/stations.txt`). I fetched it with `python3 weather/meteoswiss.py "Genève" 2026-11-05 2026-11-16T07:00` from `~/aimsight-data`, into `work/weather_geneve.csv` (05/11 → 16/11 07:00, for the scripts). Since the instructions now require it, the weather of the exact range is also in `work/weather.csv` (`python3 weather/meteoswiss.py "Genève" 2026-11-06T08:05 2026-11-16T06:27`; stderr: station GVE Genève / Cointrin, 4.9 km, 411 m; 238 hours). It is the weather a few km away; the images prevail. What it changed:
1. The housing temperature follows the **outside air** (r = +0.82, no lag) and not the sun (r = +0.02 with radiation, −0.11 with sunshine). Under the bise (13–14/11) the housing cools more than the air, and the shift of the nearest prism follows the housing then too (r = +0.96 with the housing, +0.83 with the air, best at lag 0 h). The moving element therefore follows the camera's own temperature within the hour, which a massive pier would not. This strengthens "thermal movement of the camera mount" for the 8 cycles.
2. (Withdrawn after the expert's review: the agitation episodes are no longer labelled, see §1.) The agitation of the far prisms follows the **cold of the air** (r = −0.63 with the air temperature, +0.05 with the wind). It hits only the long, low sight lines over the river. I now label it apart, as 2 **Visibility** episodes (turbulence over the warmer water, "consistent with"). The one-measurement jump of target_12/13 at 14/11 21:00 is the strongest moment of the bise episode and is named in it; the one of 07/11 05:00 has its own Displacement label again since the 07/11 morning episode was removed.
3. The two dawn transients (08/11, 14/11) fall at the coldest air of a clear night, at sunrise, inside the agitation. A bending of the long sight lines over the river now competes with a fast camera movement. Neither is established, so I moved them from Thermal displacement to **Displacement** (cause not established, both hypotheses stated).
4. The occultation of target_14 (11/11) covers the rain of that morning and ends soon after it stops: water on the prism is likely, and the description says so. The 11/11 07:15 jump falls at the start of the rain, which is now cited as possible. For the 11/11 14:45 passage the rising gusts were checked: the nearer prisms do not move in the raw image, which rules out a camera shake.
**The key to the hint (unchanged).** RD has no spot of its own. Its position falls on a dark part of the nearest pier, next to target_01's prism (`fig_rd_target01.png`). RD and target_01 have the same `error_value` at every instant and the same raw movement (`overview.py --rigid`), so RD is a pattern on target_01's prism: twins, the nearest prism. target_12 and target_13 are the farthest prisms. With temperature the nearest prism moves sideways against all the others, and the similarity fitted on RD + target_12/13 spreads this over the image. target_14, target_15 and target_02, which lie in the image between RD and the far references, take it in their corr_x. target_01, being RD's own prism, does not.
## 1. My labels
### Thermal cycles: 06/11 09:15 → 07/11 07:30; 07/11 07:30 → 08/11 07:00; 08/11 07:00 → 09/11 09:45; 09/11 09:45 → 10/11 04:45; 10/11 04:45 → 13/11 09:30; 13/11 09:30 → 14/11 07:00; 14/11 07:00 → 15/11 04:15; 15/11 04:15 → 16/11 06:15
- **Phenomenon of the catalogue**: "The camera drifts away when the temperature of its housing drops, without a daily cycle" / "The camera moves through a thermal effect". The poor compensation is a new variant: the reference network mixes the nearest and the farthest prisms, so a small movement of the mount becomes a parallax that passes into the targets between them.
- **Precedents**: CLARIDEN-5, -13, -27, -43 (reviewed): a sheltered station without a daily cycle, cut at the turning points of the housing temperature. NORDEND-83/88 (reviewed): "thermal cycle of the camera, poorly compensated", outward and return in one label, labels following each other without rest.
- **Family**: Thermal displacement. The shift follows the housing temperature and comes back with it at equal temperature: dx_RD +0.47…+0.58 at 26.8–27 °C on 07/11, 09/11, 10/11 and 11/11, and −0.05…+0.05 near 24.4 °C on 06/11, 08/11 and 13/11, with up to ~0.3 px of hysteresis. It follows the housing better than the air, at lag 0, and the housing follows the outside air, not the sun.
- **Measured**: `metrics.py` 05/11 → 16/11: r(temperature, cam_scale) = +0.97, r(cam_pan) = −0.94, r(tgt_corr_y) = +0.84. `overview.py --rigid 2026-11-06T09:00Z`: RD and target_01 dx −0.86…+0.73 px. target_02, target_14 and target_15 move less than ±0.41, and the far targets less than ±0.3. `scripts/table.py`: corr_x of target_14 from +0.69 (14/11 06:00) to −0.66 (11/11 12:00), target_01 within ±0.16. Turning points at 15 min: 06/11 09:15, 07/11 07:30, 08/11 07:00, 09/11 09:45, 10/11 04:45, 13/11 09:30, 14/11 07:00, 15/11 04:15. 10/11 → 13/11 stays one label: a plateau on 11/11 and a flat stop on 12/11, with no change of direction. `scripts/weather.py`: r(dx_RD, housing) = +0.96, r(dx_RD, air) = +0.83, r(housing, air) = +0.82, r(housing, radiation) = +0.02. The lag of dx_RD on the housing is best at 0 h. Weather of each cycle (MeteoSwiss): grey and mild 06/11; sunny midday then a clear calm night near 0 °C 07–08/11; sunny morning 08/11; mild, overcast, rain at times 10–12/11; bise (NE, gusts to ~50 km/h, dry air) from the night of 12–13/11 to 14/11; SW wind and mild air 15/11.
- **Images**: 06/11 12:00 whole image (`fig_scene_20261106T1200Z.jpg`); zone thumbnails 06/11 12:00, 14/11 03:00 (`fig_rd_target01.png`); far piers 07/11 07:00, 08/11 06:00/07:00/08:00, 11/11 07:00/08:00/14:00/15:00: nothing in front.
- **Confidence**: high on the family and on what moves; medium on the exact cut points (±1 measurement) and on the mechanism (camera mount most likely).
### Visibility: none (all removed, according to the expert)
- I had labelled four episodes of agitation of the far prisms in cold air (06/11 09:45 → 15:00; 07/11 02:45 → 10:00; 07/11 20:45 → 08/11 13:00; 13/11 02:45 → 15/11 14:00). The expert sees nothing in the data or the images; all four are removed. The statistics behind them stay in `scripts/agitation.py` (hourly median of |corr_y step| of target_07…11 a few hundredths of a pixel higher in cold air, r = −0.63 with the air temperature, nothing with the wind). They are a correlation, not a disturbance seen on the curves.
- The two one-measurement jumps of target_12/13 that I had merged into these episodes have their own Displacement labels again: 07/11 04:45 → 05:15 and 14/11 20:45 → 21:15.
### Displacement: 08/11 06:45 → 08:30 and 14/11 07:00 → 08:30 (dawn transients)
- **Phenomenon of the catalogue**: none fits exactly. Close to "A step of the camera that the compensation does not erase", but this one returns.
- **Precedents**: none with this signature.
- **Family**: Displacement (cause not established). Both come at dawn after the two coldest nights, at the coldest air hour, and end as the sun comes out (sunshine from the 08:00 hour at Cointrin both days). Two hypotheses remain: (a) a lateral bending of the long sight lines through the cold air over the water, since all the far prisms shift together and the near one does not; (b) a fast camera movement as the housing warms abruptly (+1.4 °C and +0.8 °C in the 07:00 hour, with no change in power or load in `metrics.py`).
- **Measured**: `scripts/far.py`: far prisms 03–13 raw x −0.51/−0.67/−0.38 at 08/11 07:15/07:30/07:45, back at 08:00. corr_x target_14 +0.36 (06:45) → −0.75 (07:15) → −0.09 (08:15) → +0.14 (08:30). On 14/11: far x +0.32 → −0.30 (07:45) → +0.06 (08:00); corr_x target_14 0.68 → −0.04 (07:45) → 0.40 (08:30).
- **Images**: 08/11 07:00 (haze over the river, sun on the facades), 14/11 07:00 (sun on the facades). No image during the shift.
- **Confidence**: high on the event, low on the cause.
### Displacement: 11/11 07:00 → 07:30
- target_12/13 jump alone, `error_value` 0.26/0.21 against ~0.08; corr_y of 03–11 −0.08…−0.18; back at the next measurement. The air is mild (11.7 °C, not a cold-air agitation hour); the rain begins in that hour. Cause not established (rain veiling the far references, or a partial passage). Images 07:00 and 08:00: nothing in front. Confidence: high on the event, low on the cause.
### Obstruction: 07/11 13:00 → 13:30, 15/11 14:00 → 14:30, 11/11 14:30 → 15:00
- "A pass-by hides the targets for the length of one measurement", deduced from the signature: one measurement, a group rejected, points jumping, all back at the next. On 11/11 14:45 everything beyond a given pier is lost (target_07–13, references included: no pose, the corr of 02–06 meaningless). The raw positions of the nearer prisms barely move (target_06 ≤ 0.45 px), which rules out a camera shake by the gusts rising then (Cointrin 17.6 → 28 km/h). A crossing of the line of sight, such as a boat on the river, fits that geometry but is not seen. On 07/11 and 15/11 there was some wind (gusts 18 and 33 km/h); nothing else moved. Confidence: medium (no image at the instants).
### No data: 10/11 08:15 → 09:00
- 08:30 and 08:45 are missing for all sensors, and ALASKA and NADELHORN miss exactly the same instants. WYOMING `last_sync` lists 08:45, BusVoltage is 14.1 V, and the images at 08:00 and 09:00 exist. Common cause outside the station, not established. Confidence: high.
### Obstruction: 11/11 09:30 → 14:15 (target_14)
- "An obstacle covers part of the spot". On the 12:00 thumbnail part of target_14's spot is darkened; it is whole at 07:00, 10:00, 13:00 and 15:00, and target_15 is always intact (`fig_target14_occultation.png`). `error_value` 0.077 → 0.252 (12:00) → 0.071 (13:45). corr_x(14) − corr_x(15) −0.32 at 12:00; corr_y −0.21 at 12:00, back by 14:15. Rain at Cointrin 07:00–12:00 (1.9 mm at its heaviest, in the 09:00 hour) and none from 13:00, matching the rise and the end: water on the prism is likely. Family Obstruction, since a single spot is hidden (`families.md`: "under a single spot, it is an obstacle"). Confidence: high on the occultation, medium on the water.
## 2. What I left without a label
- 06/11 09:45–15:00: a slight agitation of the far targets (hourly |corr_y step| 0.025–0.040 at 10:00, 11:00, 14:00, against 0.010–0.020 around it), air 8–10 °C. Labelled Visibility in my second version, removed: according to the expert, nothing shows in the data or the images. Too weak to be told from the ordinary noise.
- 14/11 10:15: target_12/13 opposite raw jumps and corr_y of 07–10 +0.10…+0.14 for one measurement; too small to stand out from the noise of these days.
- The agitation of the far targets in cold air (07–08/11, 13–15/11): a few hundredths of a pixel more from one measurement to the next; not labelled, according to the expert.
- 09/11 night (air 8.1–8.8 °C): agitation of the far targets 0.02–0.03, just above the mild-day level. Below what I bounded as episodes; borderline.
- 10/11 00:00–09:00: the raw y steps of target_12/13 rise during the night rain (up to 0.18 at 03:00) while the far targets' corr stay calm; nothing on the curves.
- `error_value` flags on target_14/15 alone (07/11, 08/11, 15/11), with corr within ±0.07; the slight rise of target_14's `error_value` from 10/11 10:45 to 11/11 08:00, without corr effect.
- The camera pose is missing from 16/11 00:30 to the end while the corr go on: an export matter.
- Tiny one-step deviations of RD/target_12/13 corr (references, own noise).
## 3. The documentation
- **Indications from the expert**: 06/11 09:45 → 15:00, no disturbance in the data or the images; 07/11 02:45 → 10:00, too sensitive; then all the Visibility agitation labels removed, the expert seeing nothing in the data. Lesson: an agitation of a few hundredths of a pixel, found by statistics (steps per hour, correlation with the air) and not visible on the curves, is not labelled. The expert also questioned the thermal cycle 06/11 09:15 → 07/11 07:30. I kept it: corr_x of target_14/15 move by 11–15 noises, the same signature as the other cycles in smaller, and `labeling-rules.md` §1 (D-117, 5-noise rule) asks for it; it awaits the expert's decision. Lesson: an agitation episode is labelled only when it stands clearly above the noise on the curves themselves, not on a correlation with the air temperature alone.
- **Indication from the user**: the weather data. It was not in the station folder, nor in `tools/GUIDE.md`, `tools/DATA.md` or the labeling knowledge, so I had not looked for it. It sits in `~/aimsight-data/weather/meteoswiss.py`, with the place of each station in `~/aimsight-data/extraction/stations.txt`, and is shown by the Data page. `tools/GUIDE.md` should get a weather tool or a line (air temperature, rain, wind, sunshine at the nearest MeteoSwiss station). `reading-rules.md` §3 says the housing temperature is "a witness for the lens and the mount, not for the structure nor the air". Here it follows the air closely (r = +0.82) except under the bise, which is worth recording.
- **What helped**: `families.md` "Thermal displacement or Displacement" (return at equal temperature); the CLARIDEN tunnel precedents for cutting without a daily rhythm; `reading-rules.md` §2 and §5 (`--rigid` first); `tools/DATA.md` (`x, y` are not the centroid), which made "RD is a pattern beside its prism" plausible; `reading-rules.md` §6 "a twin of a reference".
- **What was missing**: a WYOMING row in `stations.md`. It should say: a bridge seen from beneath its deck along a river, in Genève; targets on the piers; RD the twin of target_01 (same prism, nearest pier); target_12/target_13 the farthest prisms; the targets between them in the image (target_02, target_14, target_15) take the thermal parallax; only `temperature_deg_c`; one image per hour. Two phenomena are missing from the catalogue: "references at very different depths, so a small movement of the mount becomes a deformation passed into the targets between them", and "agitation of the long sight lines over water when the air is colder than the water, without the `error_value` rising".
- **Contradicted or unclear**: the catalogue line "Atmospheric turbulence… `error_value` up", and the boundary "without the `error_value` rising, an agitation… is the camera heating". Here the agitation is in the air (it follows the cold air, only on the long sight lines over the water) with a flat `error_value`. I chose Visibility against the letter of that boundary and said so in each label.
- The hint "parallax does not fit target_01": target_01 is RD's own prism, so it moves with the reference and cannot show the parallax that RD passes into the others.
- **Instructions changed during the session** (the weather fetched every time; labels detailed for retrieval; a short report in plain words on an imposed outline, with a GIF when the cause is seen over time). Labels kept as they were (already detailed); report rewritten on the new outline, with a GIF of target_14 on 11/11 07:00 → 15:00 (`target14_occultation_20261111.gif`, zone 1990,725,300,80) and the main figure redrawn in plain words.
## Difficulties
- `tools/GUIDE.md` §Conventions limits Bash to the tools without scripts or loops, while the request allows scripts; I followed the request.
- The weather source was not documented in my instructions at the start; the user pointed to it, and the instructions now require it (section "The weather").
- Hourly images only, no image quality metrics: the brief events have no image at their instant.
## Sources read
`instructions/analyst.md`; `tools/GUIDE.md`; `tools/DATA.md`; `labeling/families.md`; `labeling/labeling-rules.md`; `labeling/reading-rules.md`; `labeling/stations.md`; precedents (`labeling/scripts/precedents.py`, three queries); `request.json`; WYOMING data through `inventory.py`, `overview.py` (default and `--rigid`), `metrics.py` (temperatures; power, current, load), `summary.py`, `ranges.py --rolling 4 --instants 2`, `thumbnails.py`, whole images 06/11 12:00, 08/11 07:00, 14/11 07:00; raw CSVs of ALASKA, NADELHORN and the other stations on 10/11 08:00–08:49; WYOMING `last_sync`, `BusVoltage`; the MeteoSwiss weather (`~/aimsight-data/weather/README.md`, `meteoswiss.py`, `extraction/stations.txt`, `app/server/station.py` for the link); my scripts `scripts/table.py`, `far.py`, `instants.py`, `weather.py`, `agitation.py`, `figure.py`, `figure_weather.py`.
Binary file not shown.

After

Width:  |  Height:  |  Size: 619 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 359 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 156 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 212 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 183 KiB

@@ -0,0 +1,18 @@
{"start": "2026-11-06T09:15Z", "end": "2026-11-07T07:30Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. The housing temperature rises from late morning with the outside air (MeteoSwiss Genève-Cointrin: a grey, mild day) and the pose follows it (scale up, pan down); in the raw image the nearest prism, that of the reference RD and of its twin target_01, drifts sideways against all the other prisms, which hardly move. The compensation carries this drift into the targets between RD and the far references target_12 and target_13: the corr_x of target_14, target_15 and target_02 go down clearly, those of the far targets a little, target_01 hardly moves since it is the prism of RD. The curves stay shifted in the evening, then come back during the night as the air and the housing cool, until the turning point of the next morning. Deduced cause: the shift follows the housing temperature within the hour and comes back with it, the housing following the outside air and not the sun; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. This is not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-07T07:30Z", "end": "2026-11-08T07:00Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. The housing temperature rises strongly in the morning with the outside air (MeteoSwiss Genève-Cointrin: a sunny spell around midday) and the pose follows it (scale up, pan down): the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and the compensation passes it into the corr_x of target_14, target_15 and target_02, which go down clearly, the far targets a little. From the afternoon the air and the housing cool steadily through a clear, calm night close to freezing; the pose and the curves go back past their level and keep going the other way: by the end of the night target_14, target_15 and target_02 are clearly above their level. The abrupt shift of the morning of the 8th is labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it at equal temperature; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-08T07:00Z", "end": "2026-11-09T09:45Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. After the cold night the housing temperature rises again quickly in the morning with the outside air (MeteoSwiss Genève-Cointrin: sunny morning) and stays high until the next morning; the pose follows it (scale up, pan down), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and the compensation passes this drift into the corr_x of target_14, target_15 and target_02, which go clearly below their level, the far targets a little. In the evening and the mild night the curves come back only partly, then stay still until the start of the next warming. The abrupt shift of the morning of the 8th is labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-09T09:45Z", "end": "2026-11-10T04:45Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. The housing temperature rises again from mid-morning with the outside air, the pose follows it (scale up, pan down) and the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms; through the compensation the corr_x of target_14, target_15 and target_02 go down clearly, the far targets a little. The housing stays warm in the afternoon and evening, then cools a little during the night and the curves come back part of the way, until the turning point of the next morning. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-10T04:45Z", "end": "2026-11-13T09:30Z", "family": "Thermal displacement", "description": "Thermal movement over several days, poorly compensated. In a mild, overcast and at times rainy spell (MeteoSwiss Genève-Cointrin), the housing temperature rises on the 10th, stays high all night, rises again on the 11th to the warmest of the range, comes down in the evening and stays flat all the 12th; then the bise sets in during the night to the 13th, with a cold dry wind, and the housing cools quickly. The pose follows it (scale and pan), and the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms; through the compensation the corr_x of target_14, target_15 and target_02 go down strongly until the afternoon of the 11th, then come back, stay still on the 12th, and go past their level in the cold morning of the 13th, where they are clearly above it. The corr_y of the far targets follow the temperature too, up while warm, down in the cold morning of the 13th, target_03 and target_04 more than the others, with their error_value rising. The gap of the 10th, the occultation of target_14 and the two jumps of the 11th are labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it at equal temperature; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-13T09:30Z", "end": "2026-11-14T07:00Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. Under the bise (MeteoSwiss Genève-Cointrin: cold dry wind from the north-east, some sun around midday), the housing temperature warms up a little in the late morning, then drops again in the evening and all night to the coldest of the range. The pose follows it (scale down, pan up), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and the compensation carries it into the corr_x of target_14, target_15 and target_02, which come back part of the way in the afternoon then rise strongly above their level all night; the corr_y of the far targets target_03 to target_11 stay below their level. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-14T07:00Z", "end": "2026-11-15T04:15Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. After the cold night the housing temperature rises in the morning with the outside air and the sun (MeteoSwiss Genève-Cointrin), stays moderate all day, then cools in the evening and the night as the wind drops and light rain begins. The pose follows it (scale and pan), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and through the compensation the corr_x of target_14, target_15 and target_02 come back part of the way toward their level in the morning, then rise again above it in the night; the corr_y of the far targets stay below their level in the day and come back in the evening. The abrupt shift of the morning and the one-measurement jump of the evening are labelled apart. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-15T04:15Z", "end": "2026-11-16T06:15Z", "family": "Thermal displacement", "description": "Thermal movement, poorly compensated. After the night rain, a mild south-west wind sets in (MeteoSwiss Genève-Cointrin) and the housing temperature rises clearly from the morning to the afternoon, then cools slowly in the evening and the night. The pose follows it (scale up, pan down), the nearest prism, that of the reference RD and of its twin target_01, drifts sideways in the image against the other prisms, and through the compensation the corr_x of target_14, target_15 and target_02 go from above their level to clearly below it in the afternoon, then come back only part of the way during the night; the corr_y of the far targets come back to their level. At the end of the range the curves are still a little shifted and the camera pose is no longer exported, while the measurements go on. Deduced cause: the shift follows the housing temperature within the hour and comes back with it; a small thermal movement of the camera mount, seen as parallax between the nearest prism and the far ones, is likely. Not a movement of the structure: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-07T04:45Z", "end": "2026-11-07T05:15Z", "family": "Displacement", "description": "One-measurement jump of the far targets target_03 to target_11, in corr_x and a little in corr_y, then back: in the raw image these targets do not move, while the far references target_12 and target_13 jump alone for this measurement, without their error_value rising and without rejection; the compensation passes their jump into the targets close to them in the image, target_14, target_15 and target_02 hardly move. No image at this instant. The air is cold and very humid, calm (MeteoSwiss Genève-Cointrin). Cause not established (a disturbance of the long line of sight to the far references over the river is possible). This is not a movement of the targets: points to ignore.", "review": "2026-11-17"}
{"start": "2026-11-07T13:00Z", "end": "2026-11-07T13:30Z", "family": "Obstruction", "description": "target_03 and target_04, on neighbouring piers, are rejected for one measurement, the point of target_03 having jumped far away, then back at the next; the other targets and the references do not move. Cause deduced from the signature of a passage in front of these two targets for the length of one measurement; no image at this instant. Points missing for one measurement: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-08T06:45Z", "end": "2026-11-08T08:30Z", "family": "Displacement", "description": "Sudden shift and return, at dawn after the coldest clear night: for three measurements all the far prisms, references target_12 and target_13 included, shift sideways together in the raw image, while the nearest prism, that of the reference RD and target_01, does not follow; the compensation cannot represent it and the corr_x of target_14, target_15, target_02 and target_01 drop sharply, the corr_y of the far targets target_03 to target_09 rise, the error_value rises on several of them without any rejection; everything is back within the hour, as the sun comes out (MeteoSwiss Genève-Cointrin). The 07:00 image shows a haze over the river and the sun on the facades, nothing in front of the targets. Cause not established: a bending of the long lines of sight through the cold air over the river at sunrise, or a fast movement of the camera as its housing warms up suddenly, are both possible; no image falls during the shift. Points not to be trusted during this hour; not a movement of the structure.", "review": "2026-11-17"}
{"start": "2026-11-10T08:15Z", "end": "2026-11-10T09:00Z", "family": "No data", "description": "Two measurements missing for all sensors. ALASKA and NADELHORN miss the same two measurements: a common cause outside the station (transmission or reception of the data), not a station failure; it is not established. The station runs: a synchronization falls in the gap, the voltage is normal, the images of the hour before and after exist (one image per hour). Nothing to do on the station.", "review": "2026-11-17"}
{"start": "2026-11-11T07:00Z", "end": "2026-11-11T07:30Z", "family": "Displacement", "description": "One-measurement jump of the corr_y of the far targets target_03 to target_11, then back: in the raw image these targets do not move, while the far references target_12 and target_13 jump alone, their error_value clearly up but far from the threshold, that of target_03 and target_11 a little up too; the compensation passes the jump of these references into the targets close to them in the image. No image at this instant, nothing in front on the images of the hour before and after. Cause not established: the rain begins at that hour (MeteoSwiss Genève-Cointrin) and could veil the far references for a moment, a passage partly in front of the far end of the bridge is possible too. Not a movement: points to ignore.", "review": "2026-11-17"}
{"start": "2026-11-11T09:30Z", "end": "2026-11-11T14:15Z", "family": "Obstruction", "description": "target_14 alone: its error_value rises gradually from mid-morning, peaks around noon and falls back in the early afternoon, with a dip in between; its corr_x goes down and its corr_y down too, more than its neighbour target_15, then both come back with the error_value. In the raw image only target_14 slides, the other prisms do not. On the 12:00 thumbnail, part of the spot of target_14 is darkened, the rest intact; before and after the spot is whole. The episode covers the rain of the morning (MeteoSwiss Genève-Cointrin) and ends soon after it stops: water on the prism, a drop or a drip from the deck, is likely, though not seen as such. Ignore the curve of target_14 over the range; not a movement.", "review": "2026-11-17"}
{"start": "2026-11-11T14:30Z", "end": "2026-11-11T15:00Z", "family": "Obstruction", "description": "For one measurement the far references target_12 and target_13 and the far targets target_07 to target_11 are rejected, their points jumping in all directions, the compensation has no reference left and the camera pose is missing; the corr of target_02 to target_06, kept, jump strongly and those of target_01 and target_14 a little: the points of this measurement make no sense. All back at the next measurement. The raw positions of the nearer prisms hardly move, which rules out a shake of the camera by the gusts rising at that hour (MeteoSwiss Genève-Cointrin); everything beyond a given pier is lost at once. Cause deduced from the signature of a passage across the line of sight, in front of the far end of the bridge, for the length of one measurement; no image at this instant, nothing in front on those of 14:00 and 15:00. Points to ignore: nothing to do.", "review": "2026-11-17"}
{"start": "2026-11-14T07:00Z", "end": "2026-11-14T08:30Z", "family": "Displacement", "description": "Sudden shift and return, at dawn after the coldest night of the bise, the same signature as on the 8th: for three measurements all the far prisms, references target_12 and target_13 included, shift sideways together in the raw image, while the nearest prism, that of the reference RD and target_01, does not follow; the corr_x of target_14, target_15, target_02 and target_01 drop sharply toward their level, then rise again part of the way when the far prisms come back, as the sun comes out (MeteoSwiss Genève-Cointrin). The 07:00 image shows the sun on the facades, nothing in front of the targets. Cause not established: a bending of the long lines of sight through the cold air over the river at sunrise, or a fast movement of the camera as its housing warms up suddenly, are both possible; no image falls during the shift. Points not to be trusted during this hour; not a movement of the structure.", "review": "2026-11-17"}
{"start": "2026-11-14T20:45Z", "end": "2026-11-14T21:15Z", "family": "Displacement", "description": "One-measurement jump of the corr_y of the targets target_02 to target_11, with a smaller one in corr_x, then back: in the raw image these targets do not move, while the far references target_12 and target_13 jump alone for this measurement, without their error_value rising; the compensation passes their jump into the targets. No image at this instant. A cold, calm night under clearing bise (MeteoSwiss Genève-Cointrin). Cause not established (a disturbance of the long line of sight to the far references over the river is possible). Not a movement: points to ignore.", "review": "2026-11-17"}
{"start": "2026-11-15T14:00Z", "end": "2026-11-15T14:30Z", "family": "Obstruction", "description": "target_03 is rejected for one measurement, its point jumping far away, and the error_value of target_04 and target_15 rises at the same instant, the corr_x of target_04 jumping a little; everything is back at the next measurement and the references do not move. Cause deduced from the signature of a passage for the length of one measurement; no image at this instant. Point missing for one measurement: nothing to do.", "review": "2026-11-17"}
@@ -0,0 +1,59 @@
# WYOMING: a thermal parallax brought in by the nearest reference
You asked why target_14, target_15 and target_02 behave strangely from 6 to 16 November, and whether it is parallax, which target_01, the closest target, does not seem to show.
## In short
Yes, it is parallax, and target_01 fits it: the reference RD is placed on target_01's own prism, so target_01 moves with the reference and shows nothing. target_14, target_15 and target_02 are off by up to almost a pixel, depending on the temperature; the far targets much less. No real movement of the structure is seen.
## What happened
**Thermal parallax, all along the range.** The camera looks along the bridge under its deck. Its three references are the nearest prism (RD, the same prism as target_01) and the two farthest (target_12, target_13). When the camera's temperature changes, the nearest prism shifts sideways in the image against all the others. That is what a small movement of the camera mount does to objects at very different distances. The correction cannot represent it: it spreads the shift of the nearest prism over the image. target_14, target_15 and target_02 sit between the references in the image, so they get most of it. Their corrected curves in x go down when it is warm (most on the 11th) and up when it is cold (8th, 13th–14th), and come back at equal temperature. The camera temperature follows the outside air, not the sun, and the shift follows the camera temperature within the hour. It is the camera mount, likely; a massive pier would lag by hours.
![thermal parallax](fig_thermal_parallax.png)
Middle: RD and target_01 (one prism) move with the temperature (top) while 02, 14 and 15 stay still in the image. Bottom: their corrected curves take the opposite of that shift; target_01 stays flat.
**Two shifts at dawn, 8 and 14 November around 07:15–08:15.** After the two coldest clear nights, all the far prisms shift sideways together for three measurements, then come back. The curves of 02, 14 and 15 jump by up to almost a pixel for an hour. The cause is unknown: either the sight lines are bent through the cold air over the river at sunrise, or the camera moves fast as it warms up.
**Brief jumps from the far references, 7 Nov 05:00, 11 Nov 07:15, 14 Nov 21:00.** target_12 and target_13 jump alone for one measurement. Through the correction, all the far targets jump with them. The cause is unknown; on the 11th the rain was just starting.
**Passages.** On 7 Nov 13:15 and 15 Nov 14:15, target_03 (and target_04 on the 7th) is lost for one measurement. On 11 Nov 14:45 everything beyond one pier is lost at once, references included, and that measurement means nothing. Likely something crossing the line of sight; it is not seen, as there is one image per hour.
**target_14 partly covered, 11 Nov 09:30–14:15.** Part of its spot goes dark around noon, during the morning rain; its curve slides and comes back when the rain stops. Water on the prism is likely.
![target_14 on 11 November](target14_occultation_20261111.gif)
Look at the left spot (target_14) at 12:00: part of it darkens; target_15, on the right, stays whole.
**Two measurements missing, 10 Nov 08:30 and 08:45.** The same minutes are missing on ALASKA and NADELHORN, so the cause is outside the station.
## What to do
On the configuration: as long as the correction rests on the nearest prism (RD) and the farthest ones, the targets in between will carry this thermal parallax. Remove RD from the references, or add a reference at the distance of target_02, target_14 and target_15. Until then, read the curves of these three against the camera temperature.
## Not established
- Camera mount or the support of the nearest prism: a look at the camera mount on site would settle it.
- The cause of the dawn shifts and of the jumps of the far references: an image at each measurement, or the water temperature, would help.
## Labels
| Start (UTC) | End (UTC) | Family | Targets | In a few words |
|---|---|---|---|---|
| 06/11 09:15 | 07/11 07:30 | Thermal displacement | 02, 14, 15 (RD/01) | small warming, back at night |
| 07/11 07:30 | 08/11 07:00 | Thermal displacement | 02, 14, 15 (RD/01) | warm midday, freezing night |
| 08/11 07:00 | 09/11 09:45 | Thermal displacement | 02, 14, 15 (RD/01) | warming after the cold night |
| 09/11 09:45 | 10/11 04:45 | Thermal displacement | 02, 14, 15 (RD/01) | daily warming, partial return |
| 10/11 04:45 | 13/11 09:30 | Thermal displacement | 02, 14, 15, 03, 04 (RD/01) | mild spell, then the bise |
| 13/11 09:30 | 14/11 07:00 | Thermal displacement | 02, 14, 15, far targets (RD/01) | bise, coldest night |
| 14/11 07:00 | 15/11 04:15 | Thermal displacement | 02, 14, 15 (RD/01) | partial warming, cooling again |
| 15/11 04:15 | 16/11 06:15 | Thermal displacement | 02, 14, 15 (RD/01) | warming, not back at the end |
| 07/11 04:45 | 07/11 05:15 | Displacement | 03–11 (via 12, 13) | far references jump alone |
| 07/11 13:00 | 07/11 13:30 | Obstruction | 03, 04 | passage, one measurement |
| 08/11 06:45 | 08/11 08:30 | Displacement | all, most 02, 14, 15 | dawn shift of the far prisms |
| 10/11 08:15 | 10/11 09:00 | No data | all | gap shared with other stations |
| 11/11 07:00 | 11/11 07:30 | Displacement | 03–11 (via 12, 13) | far references jump, rain starting |
| 11/11 09:30 | 11/11 14:15 | Obstruction | 14 | spot partly covered, rain |
| 11/11 14:30 | 11/11 15:00 | Obstruction | 07–13, all through the correction | passage at the far end |
| 14/11 07:00 | 14/11 08:30 | Displacement | all, most 02, 14, 15 | dawn shift of the far prisms |
| 14/11 20:45 | 14/11 21:15 | Displacement | 02–11 (via 12, 13) | far references jump alone |
| 15/11 14:00 | 15/11 14:30 | Obstruction | 03 | passage, one measurement |
@@ -0,0 +1,6 @@
{
"station": "WYOMING",
"start": "2026-11-06T08:05Z",
"end": "2026-11-16T06:27Z",
"hints": "The behaviour of target 14, 15 and 02 is weird. I was thinking about a parallax motion but it does not fit with target 01 that is the closest"
}
@@ -0,0 +1,43 @@
"""Per hour: agitation of the far group (median over target_07..target_11 of |corr_y step| between consecutive
measurements) and of the far references (|raw y step| of target_12, target_13), against MeteoSwiss wind, humidity,
air temperature; then correlations. Also the lag between dx_RD and the housing temperature.
usage: python3 agitation.py
"""
import csv, statistics, sys
sys.path.insert(0, "tools/scripts")
import data
D = "labeling/auto/analyses/WYOMING_2026-11-06T0805_2026-11-16T0627/work/"
start, end = data.parse_time("2026-11-06T08:15Z"), data.parse_time("2026-11-16T06:15Z")
meas = data.load("data/WYOMING")
w = {data.parse_time(r["time_utc"]): r for r in csv.DictReader(open(D + "weather_geneve.csv"))}
FAR = ["target_%02d" % i for i in range(7, 12)]
hours = {}
for s in FAR + ["target_12", "target_13"]:
v = [m for m in meas[s] if start <= m["t"] <= end]
for a, b in zip(v, v[1:]):
h = b["t"].replace(minute=0, second=0, microsecond=0)
if s in FAR:
if a["corr_y"] is not None and b["corr_y"] is not None:
hours.setdefault(h, {}).setdefault("tgt", []).append(abs(b["corr_y"] - a["corr_y"]))
elif not data.rejected(a) and not data.rejected(b):
hours.setdefault(h, {}).setdefault("ref", []).append(abs(b["y"] - a["y"]))
rows = []
for h in sorted(hours):
if h in w and "tgt" in hours[h] and "ref" in hours[h]:
f = lambda k: float(w[h][k]) if w[h][k] else None
rows.append((h, statistics.median(hours[h]["tgt"]), statistics.median(hours[h]["ref"]), f("wind_speed_kmh"),
f("wind_gust_kmh"), f("humidity_pct"), f("temperature_c"), f("global_radiation_wm2")))
by_day = {}
for r in rows:
by_day.setdefault(r[0].strftime("%m-%d"), []).append(r)
print("day agit_tgt agit_ref wind gust (daily medians)")
for d, rs in by_day.items():
print(d, "%.3f %.3f %.1f %.1f" % tuple(statistics.median([x[i] for x in rs]) for i in (1, 2, 3, 4)))
def corr(i, j):
p = [(r[i], r[j]) for r in rows if r[i] is not None and r[j] is not None]
return statistics.correlation(*zip(*p))
for name, j in (("wind", 3), ("gust", 4), ("humidity", 5), ("air", 6), ("radiation", 7)):
print(f"r(agit_tgt, {name}) = {corr(1, j):+.2f}; r(agit_ref, {name}) = {corr(2, j):+.2f}")
print(f"r(agit_tgt, agit_ref) = {corr(1, 2):+.2f}; n = {len(rows)}")
@@ -0,0 +1,25 @@
"""Per 15-min measurement: median raw x and y shift of the far group (target_03..target_13) from its window median,
RD and target_01 raw x shift, and corr_x of target_14, between hours H1 and H2 of each day.
usage: python3 far.py START END H1 H2
"""
import statistics, sys
sys.path.insert(0, "tools/scripts")
import data
F = "data/WYOMING"
start, end = data.parse_time(sys.argv[1]), data.parse_time(sys.argv[2])
h1, h2 = int(sys.argv[3]), int(sys.argv[4])
meas = data.load(F)
FAR = ["target_%02d" % i for i in range(3, 14)]
by = {s: {m["t"].replace(second=0, microsecond=0): m for m in v} for s, v in meas.items()}
win = {s: [m for m in v if start <= m["t"] <= end] for s, v in meas.items()}
base = {s: (statistics.median([m["x"] for m in v]), statistics.median([m["y"] for m in v])) for s, v in win.items()}
for t in sorted(by["RD"]):
if not (start <= t <= end and h1 <= t.hour < h2):
continue
fx = statistics.median([by[s][t]["x"] - base[s][0] for s in FAR if t in by[s]])
fy = statistics.median([by[s][t]["y"] - base[s][1] for s in FAR if t in by[s]])
rd = by["RD"][t]["x"] - base["RD"][0]
c14 = by["target_14"][t]["corr_x"]
print(t.strftime("%m-%dT%H:%M"), "far_dx %+.2f far_dy %+.2f RD_dx %+.2f c14 %s" % (fx, fy, rd, "-" if c14 is None else "%+.2f" % c14))
@@ -0,0 +1,30 @@
"""Figure for the report: housing temperature, raw sideways shift of the nearest prism (RD, target_01) against the far
prisms, and corr_x of target_02, target_14, target_15, target_01, hourly medians from table.py's output.
usage: .venv/bin/python figure.py TABLE OUT.png
"""
import sys
from datetime import datetime
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
rows = [l.rstrip("\n").split("\t") for l in open(sys.argv[1])]
head, rows = rows[0], [r for r in rows[1:] if r[0] >= "11-06T08"]
col = {h: i for i, h in enumerate(head)}
t = [datetime.strptime("2026-" + r[0], "%Y-%m-%dT%H:%M") for r in rows]
def v(name):
return [float(r[col[name]]) if r[col[name]] != "-" else None for r in rows]
fig, ax = plt.subplots(3, 1, figsize=(13, 9), sharex=True)
ax[0].plot(t, v("temp"), "k"); ax[0].set_ylabel("camera housing\ntemperature [°C]")
ax[1].plot(t, v("dx_RD"), label="RD (reference, same prism as target_01)"); ax[1].plot(t, v("dx_01"), label="target_01")
for s in ("02", "14", "15"):
ax[1].plot(t, v("dx_" + s), label="target_" + s)
ax[1].set_ylabel("sideways shift in the image,\nagainst the far prisms [px]"); ax[1].legend(fontsize=8)
for s in ("02", "14", "15", "01"):
ax[2].plot(t, v("cx" + s), label="target_" + s)
ax[2].set_ylabel("corrected curve in x [px]"); ax[2].legend(fontsize=8)
for a in ax:
a.grid(alpha=.3)
fig.suptitle("WYOMING: the nearest prism shifts with the temperature, and the targets between the references take the shift (hourly, UTC)")
fig.tight_layout(); fig.savefig(sys.argv[2], dpi=90); print(sys.argv[2])
@@ -0,0 +1,35 @@
"""Report figure: MeteoSwiss Genève-Cointrin air temperature, wind, sunshine against the housing temperature, the
nearest prism's shift (dx_RD) and the hourly agitation of the far targets (from agitation.py's computation).
usage: .venv/bin/python figure_weather.py OUT.png
"""
import csv, statistics, sys
sys.path.insert(0, "tools/scripts")
import data
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
D = "labeling/auto/analyses/WYOMING_2026-11-06T0805_2026-11-16T0627/"
src = open(D + "scripts/agitation.py").read().split("by_day = {}")[0]
ns = {}
exec(src, ns)
rows = ns["rows"]
t = [r[0] for r in rows]
tab = {l.split("\t")[0]: l.rstrip("\n").split("\t") for l in open(D + "work/table_1h.txt")}
house = [float(tab[x.strftime("%m-%dT%H:%M")][1]) for x in t]
dx = [float(tab[x.strftime("%m-%dT%H:%M")][4]) for x in t]
w = {data.parse_time(r["time_utc"]): r for r in csv.DictReader(open(D + "work/weather_geneve.csv"))}
sun = [float(w[x]["sunshine_min"] or 0) for x in t]
fig, ax = plt.subplots(4, 1, figsize=(13, 10), sharex=True)
ax[0].plot(t, house, "k", label="housing temperature"); a2 = ax[0].twinx()
a2.plot(t, [r[6] for r in rows], "tab:red", label="air, Genève-Cointrin"); ax[0].set_ylabel("housing [°C]"); a2.set_ylabel("air [°C]", color="tab:red")
ax[1].plot(t, dx, "tab:blue"); ax[1].set_ylabel("RD = target_01 raw x\nminus far prisms [px]")
ax[2].plot(t, [r[1] for r in rows], "tab:purple", label="far targets 07-11, |corr_y step|")
ax[2].plot(t, [r[2] for r in rows], "tab:orange", alpha=.6, label="refs 12/13, |raw y step|"); ax[2].set_ylabel("agitation [px]"); ax[2].legend(fontsize=8)
ax[3].plot(t, [r[3] for r in rows], "tab:green", label="wind [km/h]"); ax[3].bar(t, sun, width=0.04, color="gold", label="sunshine [min/h]")
ax[3].set_ylabel("wind, sun"); ax[3].legend(fontsize=8)
for a in ax:
a.grid(alpha=.3)
fig.suptitle("WYOMING against the MeteoSwiss weather of Genève-Cointrin (hourly, UTC)")
fig.tight_layout(); fig.savefig(sys.argv[1], dpi=90); print(sys.argv[1])
@@ -0,0 +1,31 @@
"""At chosen instants (and the measurement before and after): per sensor, corr_x and corr_y minus the median of the
4 measurements on each side (without the instants +-1), error_value; R = rejected.
usage: python3 instants.py T1,T2,...
"""
import statistics, sys
from datetime import timedelta
sys.path.insert(0, "tools/scripts")
import data
meas = data.load("data/WYOMING")
by = {s: {m["t"].replace(second=0, microsecond=0): m for m in v} for s, v in meas.items()}
q = timedelta(minutes=15)
for txt in sys.argv[1].split(","):
t0 = data.parse_time(txt)
print("=== around", txt)
for t in (t0 - q, t0, t0 + q):
cells = []
for s in sorted(by):
if t not in by[s]:
cells.append(s[-2:] + ":-"); continue
m = by[s][t]
if data.rejected(m):
cells.append(s[-2:] + ":R"); continue
nb = [by[s].get(t0 + k * q) for k in list(range(-6, -1)) + list(range(2, 7))]
nb = [n for n in nb if n and n["corr_x"] is not None]
dx = m["corr_x"] - statistics.median([n["corr_x"] for n in nb])
dy = m["corr_y"] - statistics.median([n["corr_y"] for n in nb])
flag = "*" if max(abs(dx), abs(dy)) > 0.08 else ""
cells.append("%s:%+.2f/%+.2f%s" % (s[-2:], dx, dy, flag))
print(t.strftime("%dT%H:%M"), " ".join(cells))
@@ -0,0 +1,48 @@
"""Per step: temperature, raw x/y of near sensors minus the far group (translation only), corr_x of chosen targets.
usage: python3 table.py START END [STEP_MIN] (UTC, from the aimsight-app root)
"""
import statistics, sys
from datetime import timedelta
sys.path.insert(0, "tools/scripts")
import data
F = "data/WYOMING"
start, end = data.parse_time(sys.argv[1]), data.parse_time(sys.argv[2])
step = int(sys.argv[3]) if len(sys.argv) > 3 else 60
meas = data.load(F)
temp = dict(data.load_temperature(F)) or {}
if not temp:
temp = dict(data.load_metrics(F).get("temperature_deg_c", []))
cam = {p["t"].replace(second=0, microsecond=0): p for p in data.load_camera(F)}
FAR = ["target_%02d" % i for i in range(3, 14)]
NEAR = ["RD", "target_01", "target_02", "target_14", "target_15"]
by = {s: {m["t"].replace(second=0, microsecond=0): m for m in v} for s, v in meas.items()}
base = {s: (statistics.median([m["x"] for m in v]), statistics.median([m["y"] for m in v])) for s, v in meas.items()}
ts = sorted(by["RD"])
print("time\ttemp\tscale_e4\tpan_e3\t" + "\t".join("dx_" + s[-2:] for s in NEAR) + "\t" + "\t".join("dy_" + s[-2:] for s in NEAR[:2]) + "\tcx02\tcx14\tcx15\tcx01\tev_RD\tev_01")
t = start
while t <= end:
sel = [u for u in ts if t <= u < t + timedelta(minutes=step)]
if sel:
row = []
def med(f):
v = [f(u) for u in sel]
v = [x for x in v if x is not None]
return statistics.median(v) if v else None
def far(u, c, k):
return statistics.median([by[s][u][c] - base[s][k] for s in FAR if u in by[s] and by[s][u]["corr_x"] is not None] or [0])
th = t.replace(minute=0)
row.append(temp.get(th))
row.append(med(lambda u: (cam[u]["scale"] - 1) * 1e4 if u in cam else None))
row.append(med(lambda u: cam[u]["pan"] * 1e3 if u in cam else None))
for s in NEAR:
row.append(med(lambda u: by[s][u]["x"] - base[s][0] - far(u, "x", 0) if u in by[s] else None))
for s in NEAR[:2]:
row.append(med(lambda u: by[s][u]["y"] - base[s][1] - far(u, "y", 1) if u in by[s] else None))
for s in ["target_02", "target_14", "target_15", "target_01"]:
row.append(med(lambda u: by[s][u]["corr_x"] if u in by[s] else None))
for s in ["RD", "target_01"]:
row.append(med(lambda u: by[s][u]["error_value"] if u in by[s] else None))
print(t.strftime("%m-%dT%H:%M") + "\t" + "\t".join("-" if v is None else "%.2f" % v for v in row))
t += timedelta(minutes=step)
@@ -0,0 +1,42 @@
"""Join the MeteoSwiss hourly weather (work/weather_geneve.csv) with table.py's hourly output (work/table_1h.txt):
prints the joined table (every N hours) and correlations of dx_RD, temp (housing) with the weather columns, plus lags.
usage: python3 weather.py [EVERY]
"""
import csv, statistics, sys
D = "labeling/auto/analyses/WYOMING_2026-11-06T0805_2026-11-16T0627/work/"
every = int(sys.argv[1]) if len(sys.argv) > 1 else 1
w = {r["time_utc"][5:16]: r for r in csv.DictReader(open(D + "weather_geneve.csv"))}
rows = [l.rstrip("\n").split("\t") for l in open(D + "table_1h.txt")]
head, rows = rows[0], rows[1:]
c = {h: i for i, h in enumerate(head)}
W = ["temperature_c", "sunshine_min", "global_radiation_wm2", "precipitation_mm", "humidity_pct", "wind_speed_kmh", "wind_direction_deg"]
joined = []
for r in rows:
k = r[0]
if k in w:
def f(x):
try: return float(x)
except ValueError: return None
joined.append((k, f(r[c["temp"]]), f(r[c["dx_RD"]]), f(r[c["cx14"]]), *[f(w[k][x]) for x in W]))
print("time housing dx_RD cx14 air sun rad rain hum wind wdir")
for i, j in enumerate(joined):
if i % every == 0 and j[0] >= "11-06T08":
print(" ".join("-" if v is None else (v if isinstance(v, str) else "%.1f" % v) for v in j))
def corr(a, b):
p = [(x, y) for x, y in zip(a, b) if x is not None and y is not None]
xs, ys = zip(*p)
return statistics.correlation(xs, ys), len(p)
sel = [j for j in joined if j[0] >= "11-06T08"]
cols = {"housing": 1, "dx_RD": 2, "air": 4, "sun": 5, "rad": 6, "hum": 8, "wind": 9}
for a in ("housing", "dx_RD"):
for b in ("air", "rad", "sun", "hum", "wind") + (("housing",) if a == "dx_RD" else ()):
r, n = corr([s[cols[a]] for s in sel], [s[cols[b]] for s in sel])
print(f"r({a}, {b}) = {r:+.2f} (n={n})")
# lag: dx_RD against air temperature shifted by L hours (air earlier)
for L in (0, 1, 2, 3, 4, 6):
a = [s[2] for s in sel[L:]]; b = [s[4] for s in sel[:len(sel) - L]]
print(f"lag {L} h: r(dx_RD, air earlier) = {corr(a, b)[0]:+.2f}; r(housing, air earlier) = {corr([s[1] for s in sel[L:]], b)[0]:+.2f}")
Binary file not shown.

After

Width:  |  Height:  |  Size: 423 KiB