Step 4: M2, multiple-instance model in PyTorch (shared sensor MLP, attention per family, averaged seeds); the time-split report shared with M1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
975ebc7c1a
commit
fef30b5ef4
4 files changed
+153
-16
No files matched your search
@@ -49,28 +49,18 @@ def fit(x, y, w=None):
|
||||
|
||||
def split_time(a, data, rows, families):
|
||||
"""Every station seen in training: does the information lie in the features, apart from moving to a new station?"""
|
||||
from evaluate import counts, regressions, show, total
|
||||
cut = {s: int(0.7 * len(r)) for s, r in rows.items()}
|
||||
from evaluate import CUT, report_time_split
|
||||
cut = {s: int(CUT * len(r)) for s, r in rows.items()}
|
||||
x = np.concatenate([rows[s][:cut[s]] for s in data])
|
||||
base, new = {}, {}
|
||||
yhat = {s: np.load(Path(a.cache) / "rules" / f"{s}.npz")["yhat"][cut[s]:].copy() for s in data}
|
||||
for f in families:
|
||||
proba = {s: np.zeros((len(r) - cut[s], len(families))) for s, r in rows.items()}
|
||||
for j, f in enumerate(families):
|
||||
k = FAMILIES.index(f)
|
||||
y = np.concatenate([data[s]["y"][:cut[s], k] for s in data])
|
||||
w = np.concatenate([np.full(cut[s], 1 / cut[s]) for s in data]) if a.balance else None
|
||||
model = fit(x, y, w)
|
||||
for s in data:
|
||||
yhat[s][:, k] = model.predict_proba(rows[s][cut[s]:])[:, 1] >= a.threshold
|
||||
for s in data:
|
||||
y = data[s]["y"][cut[s]:]
|
||||
base[s] = counts(y, np.load(Path(a.cache) / "rules" / f"{s}.npz")["yhat"][cut[s]:])
|
||||
new[s] = counts(y, yhat[s])
|
||||
keep = lambda c: {f: v for f, v in c.items() if f in families} # noqa: E731
|
||||
for s in data:
|
||||
for f, name, pb, pn in regressions(keep(base[s]), keep(new[s])):
|
||||
print(f" regression: {s}, {f}, {name} {pb:.0f} % -> {'—' if pn is None else f'{pn:.0f} %'}"
|
||||
f" (rules {base[s][f]}, model {new[s].get(f)}: labelled, flagged, both)")
|
||||
show("Last 30 % of every station", keep(total(base)), keep(total(new)))
|
||||
proba[s][:, j] = model.predict_proba(rows[s][cut[s]:])[:, 1]
|
||||
report_time_split(a.cache, data, proba, families, a.threshold)
|
||||
|
||||
|
||||
def main():
|
||||
|
||||
Reference in new issue
Block a user