import inspect
from importlib import import_module
import econml
import numpy as np
import pandas as pd
from dowhy.causal_estimator import CausalEstimate, CausalEstimator
from dowhy.utils.api import parse_state
[docs]class Econml(CausalEstimator):
"""Wrapper class for estimators from the EconML library.
For a list of standard args and kwargs, see documentation for
:class:`~dowhy.causal_estimator.CausalEstimator`.
Supports additional parameters as listed below. For init and fit
parameters of each estimator, refer to the EconML docs.
"""
def __init__(self, *args, econml_methodname, **kwargs):
"""
:param econml_methodname: Fully qualified name of econml estimator
class. For example, 'econml.dml.DML'
"""
# Required to ensure that self.method_params contains all the
# parameters to create an object of this class
args_dict = {k: v for k, v in locals().items() if k not in type(self)._STD_INIT_ARGS}
args_dict.update(kwargs)
super().__init__(*args, **args_dict)
self._econml_methodname = econml_methodname
self.logger.info("INFO: Using EconML Estimator")
self.identifier_method = self._target_estimand.identifier_method
self._observed_common_causes_names = self._target_estimand.get_backdoor_variables().copy()
# For metalearners only--issue a warning if w contains variables not in x
(module_name, _, class_name) = self._econml_methodname.rpartition(".")
if module_name.endswith("metalearners"):
effect_modifier_names = []
if self._effect_modifier_names is not None:
effect_modifier_names = self._effect_modifier_names.copy()
w_diff_x = [w for w in self._observed_common_causes_names if w not in effect_modifier_names]
if len(w_diff_x) > 0:
self.logger.warn(
"Concatenating common_causes and effect_modifiers and providing a single list of variables to metalearner estimator method, "
+ class_name
+ ". EconML metalearners accept a single X argument."
)
effect_modifier_names.extend(w_diff_x)
# Override the effect_modifiers set in CausalEstimator.__init__()
# Also only update self._effect_modifiers, and create a copy of self._effect_modifier_names
# the latter can be used by other estimator methods later
self._effect_modifiers = self._data[effect_modifier_names]
self._effect_modifiers = pd.get_dummies(self._effect_modifiers, drop_first=True)
self._effect_modifier_names = effect_modifier_names
self.logger.debug("Effect modifiers: " + ",".join(effect_modifier_names))
if self._observed_common_causes_names:
self._observed_common_causes = self._data[self._observed_common_causes_names]
self._observed_common_causes = pd.get_dummies(self._observed_common_causes, drop_first=True)
else:
self._observed_common_causes = None
self.logger.debug("Back-door variables used:" + ",".join(self._observed_common_causes_names))
# Instrumental variables names, if present
# choosing the instrumental variable to use
if getattr(self, "iv_instrument_name", None) is None:
self.estimating_instrument_names = self._target_estimand.instrumental_variables
else:
self.estimating_instrument_names = parse_state(self.iv_instrument_name)
if self.estimating_instrument_names:
self._estimating_instruments = self._data[self.estimating_instrument_names]
self._estimating_instruments = pd.get_dummies(self._estimating_instruments, drop_first=True)
else:
self._estimating_instruments = None
self.estimator = None
self.symbolic_estimator = self.construct_symbolic_estimator(self._target_estimand)
self.logger.info(self.symbolic_estimator)
def _get_econml_class_object(self, module_method_name, *args, **kwargs):
# from https://www.bnmetrics.com/blog/factory-pattern-in-python3-simple-version
try:
(module_name, _, class_name) = module_method_name.rpartition(".")
estimator_module = import_module(module_name)
estimator_class = getattr(estimator_module, class_name)
except (AttributeError, AssertionError, ImportError):
raise ImportError(
"Error loading {}.{}. Double-check the method name and ensure that all econml dependencies are installed.".format(
module_name, class_name
)
)
return estimator_class
def _estimate_effect(self):
n_samples = self._treatment.shape[0]
X = None # Effect modifiers
W = None # common causes/ confounders
Z = None # Instruments
Y = self._outcome
T = self._treatment
if self._effect_modifiers is not None:
X = self._effect_modifiers
if self._observed_common_causes_names:
W = self._observed_common_causes
if self.estimating_instrument_names:
Z = self._estimating_instruments
named_data_args = {"Y": Y, "T": T, "X": X, "W": W, "Z": Z}
if self.estimator is None:
estimator_class = self._get_econml_class_object(self._econml_methodname)
self.estimator = estimator_class(**self.method_params["init_params"])
# Calling the econml estimator's fit method
estimator_argspec = inspect.getfullargspec(inspect.unwrap(self.estimator.fit))
# As of v0.9, econml has some kewyord only arguments
estimator_named_args = estimator_argspec.args + estimator_argspec.kwonlyargs
estimator_data_args = {
arg: named_data_args[arg] for arg in named_data_args.keys() if arg in estimator_named_args
}
if self.method_params["fit_params"] is not False:
self.estimator.fit(**estimator_data_args, **self.method_params["fit_params"])
X_test = X
n_target_units = n_samples
if X is not None:
if type(self._target_units) is pd.DataFrame:
X_test = self._target_units
elif callable(self._target_units):
filtered_rows = self._data.where(self._target_units)
boolean_criterion = np.array(filtered_rows.notnull().iloc[:, 0])
X_test = X[boolean_criterion]
n_target_units = X_test.shape[0]
# Changing shape to a list for a singleton value
if type(self._control_value) is not list:
self._control_value = [self._control_value]
if type(self._treatment_value) is not list:
self._treatment_value = [self._treatment_value]
T0_test = np.repeat([self._control_value], n_target_units, axis=0)
T1_test = np.repeat([self._treatment_value], n_target_units, axis=0)
est = self.estimator.effect(X_test, T0=T0_test, T1=T1_test)
ate = np.mean(est)
self.effect_intervals = None
if self._confidence_intervals:
self.effect_intervals = self.estimator.effect_interval(
X_test, T0=T0_test, T1=T1_test, alpha=1 - self.confidence_level
)
estimate = CausalEstimate(
estimate=ate,
control_value=self._control_value,
treatment_value=self._treatment_value,
target_estimand=self._target_estimand,
realized_estimand_expr=self.symbolic_estimator,
cate_estimates=est,
effect_intervals=self.effect_intervals,
_estimator_object=self.estimator,
)
return estimate
def _estimate_confidence_intervals(self, confidence_level=None, method=None):
"""Returns None if the confidence interval has not been calculated."""
return self.effect_intervals
def _do(self, x):
raise NotImplementedError
[docs] def construct_symbolic_estimator(self, estimand):
expr = "b: " + ", ".join(estimand.outcome_variable) + "~"
# TODO -- fix: we are actually conditioning on positive treatment (d=1)
(module_name, _, class_name) = self._econml_methodname.rpartition(".")
if module_name.endswith("metalearners"):
var_list = estimand.treatment_variable + self._effect_modifier_names
expr += "+".join(var_list)
else:
var_list = estimand.treatment_variable + self._observed_common_causes_names
expr += "+".join(var_list)
expr += " | " + ",".join(self._effect_modifier_names)
return expr
[docs] def shap_values(self, df: pd.DataFrame, *args, **kwargs):
return self.estimator.shap_values(df[self._effect_modifier_names].values, *args, **kwargs)
[docs] def effect(self, df: pd.DataFrame, *args, **kwargs) -> np.ndarray:
return self.estimator.effect(df[self._effect_modifier_names].values, *args, **kwargs)
[docs] def effect_inference(self, df: pd.DataFrame, *args, **kwargs) -> np.ndarray:
return self.estimator.effect_inference(df[self._effect_modifier_names].values, *args, **kwargs)