def test_LearnerSCCS_fit_KFold_CV(self):
lrn = ConvSCCS(n_lags=self.n_lags,
penalized_features=np.arange(self.n_features),
random_state=self.seed,
C_tv=1e-1,
C_group_l1=1e-1)
lrn.fit(self.features, self.labels, self.censoring)
score = lrn.score()
tv_range = (-5, -1)
groupl1_range = (-5, -1)
lrn.fit_kfold_cv(self.features,
self.labels,
self.censoring,
C_tv_range=tv_range,
C_group_l1_range=groupl1_range,
n_cv_iter=4)
self.assertTrue(lrn.score() <= score)
session.merge(sim_log)
session.commit()
n_features = n_base_features
n_lags = np.repeat(49, n_features).astype("uint64")
start = time()
lrn = ConvSCCS(
n_lags=n_lags, penalized_features=np.arange(n_features), verbose=False
)
C_tv_range = (1, 5)
C_group_l1_range = (1, 5)
fitted_coeffs, cv_track = lrn.fit_kfold_cv(
censored_features,
labels,
censoring,
C_tv_range=C_tv_range,
C_group_l1_range=C_group_l1_range,
confidence_intervals=False,
) # WARNING: no bootstrap in this simulation
elapsed_time = time() - start
model_id = "ConvSCCS"
model_log = ConvSCCSModel(
experiment_id=experiment_id,
version=version,
seed=seed,
model_id=model_id,
run_time=elapsed_time,
model_params=str(cv_track.model_params),
cv_track=dumps(cv_track),
)
features = [hstack([f, feat_agegrp]).tocsr() for f in features]
censored_features = [
hstack([f, feat_agegrp]).tocsr() for f in censored_features
]
n_lags = np.hstack([n_lags, np.zeros(n_agegrps)])
start = time()
lrn = ConvSCCS(
n_lags=n_lags, penalized_features=np.arange(n_features), verbose=False
)
C_tv_range = (1, 5)
C_group_l1_range = (1, 5)
fitted_coeffs, cv_track = lrn.fit_kfold_cv(
censored_features,
labels,
censoring,
C_tv_range=C_tv_range,
C_group_l1_range=C_group_l1_range,
confidence_intervals=True
)
elapsed_time = time() - start
model_id = "ConvSCCS"
model_log = ConvSCCSModel(
experiment_id=experiment_id,
version=version,
seed=seed,
model_id=model_id,
run_time=elapsed_time,
model_params=str(cv_track.model_params),
cv_track=dumps(cv_track),
)
