Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Reuse conditional average outcome estimates for X-Learner pseudo outcome #83

Merged
merged 1 commit into from
Aug 15, 2024
Merged

Reuse conditional average outcome estimates for X-Learner pseudo outcome #83

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
65 changes: 37 additions & 28 deletions metalearners/xlearner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -173,9 +173,17 @@ def fit_all_treatment(
qualified_fit_params = self._qualified_fit_params(fit_params)

treatment_jobs: list[_ParallelJoblibSpecification] = []

conditional_average_outcome_estimates = (
self.predict_conditional_average_outcomes(
X=X,
is_oos=False,
)
)

for treatment_variant in range(1, self.n_variants):
imputed_te_control, imputed_te_treatment = self._pseudo_outcome(
X, y, w, treatment_variant
y, w, treatment_variant, conditional_average_outcome_estimates
)
treatment_jobs.append(
self._treatment_joblib_specifications(
Expand Down Expand Up @@ -270,6 +278,7 @@ def predict(
oos_method=oos_method,
)
)

tau_hat_treatment[treatment_variant_indices] = self.predict_treatment(
X=index_matrix(X, treatment_variant_indices),
model_kind=TREATMENT_EFFECT_MODEL,
Expand Down Expand Up @@ -350,11 +359,19 @@ def evaluate(
feature_set=self.feature_set[PROPENSITY_MODEL],
)

conditional_average_outcome_estimates = (
self.predict_conditional_average_outcomes(
X=X,
is_oos=is_oos,
oos_method=oos_method,
)
)

imputed_te_control: list[np.ndarray] = []
imputed_te_treatment: list[np.ndarray] = []
for treatment_variant in range(1, self.n_variants):
tv_imputed_te_control, tv_imputed_te_treatment = self._pseudo_outcome(
X, y, w, treatment_variant
y, w, treatment_variant, conditional_average_outcome_estimates
)
imputed_te_control.append(tv_imputed_te_control)
imputed_te_treatment.append(tv_imputed_te_treatment)
Expand Down Expand Up @@ -391,7 +408,11 @@ def evaluate(
)

def _pseudo_outcome(
self, X: Matrix, y: Vector, w: Vector, treatment_variant: int
self,
y: Vector,
w: Vector,
treatment_variant: int,
conditional_average_outcome_estimates: np.ndarray,
) -> tuple[np.ndarray, np.ndarray]:
"""Compute the X-Learner pseudo outcome.

Expand All @@ -404,38 +425,26 @@ def _pseudo_outcome(
"""
validate_valid_treatment_variant_not_control(treatment_variant, self.n_variants)

mask = (w == treatment_variant) | (w == 0)
treatment_indices = w == treatment_variant
control_indices = w == 0

X_filt = X[mask]
y_filt = y[mask]
w_filt = (w[mask] == treatment_variant).astype(int)
treatment_outcome = index_matrix(
conditional_average_outcome_estimates, control_indices
)[:, treatment_variant]
control_outcome = index_matrix(
conditional_average_outcome_estimates, treatment_indices
)[:, 0]

treatment_indices = w_filt == 1
control_indices = w_filt == 0

# This is always oos because the VARIANT_OUTCOME_MODEL[0] is used to predict the
# control outcomes of the treated observations and vice versa.
control_outcome = self.predict_nuisance(
X=index_matrix(X_filt, treatment_indices),
model_kind=VARIANT_OUTCOME_MODEL,
model_ord=0,
is_oos=True,
oos_method=OVERALL,
)
treatment_outcome = self.predict_nuisance(
X=index_matrix(X_filt, control_indices),
model_kind=VARIANT_OUTCOME_MODEL,
model_ord=treatment_variant,
is_oos=True,
oos_method=OVERALL,
)
if self.is_classification:
# Get the probability of positive class, multiclass is currently not supported.
control_outcome = control_outcome[:, 1]
treatment_outcome = treatment_outcome[:, 1]
else:
control_outcome = control_outcome[:, 0]
treatment_outcome = treatment_outcome[:, 0]

imputed_te_treatment = y_filt[treatment_indices] - control_outcome
imputed_te_control = treatment_outcome - y_filt[control_indices]
imputed_te_treatment = y[treatment_indices] - control_outcome
imputed_te_control = treatment_outcome - y[control_indices]

return imputed_te_control, imputed_te_treatment

Expand Down
Loading