dowhy: `distance_matching` estimator raises an exception when binary treatment encoded using `int`s

Describe the bug When using distance_matching estimator with binary treatment encoded as int type (1s and 0s rather than True and False), DoWhy throws the following exception:

---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
Input In [418], in <cell line: 2>()
      1 # Get estimate (Linear Regression)
----> 2 estimate = model.estimate_effect(
      3     identified_estimand=estimand,
      4     method_name="backdoor.distance_matching",
      5     target_units="ate",
      6     method_params={'distance_metric': "minkowski", 'p':2})

File ~\anaconda3\envs\causal_book_py38\lib\site-packages\dowhy\causal_model.py:297, in CausalModel.estimate_effect(self, identified_estimand, method_name, control_value, treatment_value, test_significance, evaluate_effect_strength, confidence_intervals, target_units, effect_modifiers, fit_estimator, method_params)
    295     if method_params is None:
    296         method_params = {}
--> 297     self.causal_estimator = causal_estimator_class(
    298         self._data,
    299         identified_estimand,
    300         self._treatment, self._outcome, #names of treatment and outcome
    301         control_value = control_value,
    302         treatment_value = treatment_value,
    303         test_significance=test_significance,
    304         evaluate_effect_strength=evaluate_effect_strength,
    305         confidence_intervals = confidence_intervals,
    306         target_units = target_units,
    307         effect_modifiers = effect_modifiers,
    308         **method_params,
    309         **extra_args)
    310 else:
    311     # Estimator had been computed in a previous call
    312     assert self.causal_estimator is not None

File ~\anaconda3\envs\causal_book_py38\lib\site-packages\dowhy\causal_estimators\distance_matching_estimator.py:45, in DistanceMatchingEstimator.__init__(self, num_matches_per_unit, distance_metric, exact_match_cols, *args, **kwargs)
     43     error_msg = "Distance Matching method is applicable only for binary treatments"
     44     self.logger.error(error_msg)
---> 45     raise Exception(error_msg)
     47 self.num_matches_per_unit = num_matches_per_unit
     48 self.distance_metric = distance_metric

Exception: Distance Matching method is applicable only for binary treatments

Note that when using the same dataset with a binary treatment encoded as bool no error occurs.

Steps to reproduce the behavior

form scipy import stats
import numpy as np
import pandas as pd
import dowhy
from dowhy import CausalModel

# Generate the data
SAMPLE_SIZE = 10_000
MAX_AGE = 50

age = stats.halfnorm.rvs(loc=19, scale=10, size=SAMPLE_SIZE).astype(int)
age = np.where(age > MAX_AGE, np.random.choice(np.arange(20, MAX_AGE)), age)

took_a_course = stats.bernoulli(p=10/age).rvs()#.astype(bool)

earnings = 75000 + took_a_course * 10000 + age * 1000 + age**2 * 50 + np.random.randn(SAMPLE_SIZE) * 2000
earnings = earnings.round()


# Construct the graph (the graph is constant for all iterations)
nodes = ['took_a_course', 'earnings', 'age']
edges = [
    ('took_a_course', 'earnings'),
    ('age', 'took_a_course'),
    ('age', 'earnings')
]

# Generate the GML graph
gml_string = 'graph [directed 1\n'

for node in nodes:
    gml_string += f'\tnode [id "{node}" label "{node}"]\n'

for edge in edges:
    gml_string += f'\tedge [source "{edge[0]}" target "{edge[1]}"]\n'
    
gml_string += ']'

data = pd.DataFrame(dict(
    age=age,
    took_a_course=took_a_course,
    earnings=earnings
))


# Instantiate the CausalModel 
model = CausalModel(
    data=data,
    treatment='took_a_course',
    outcome='earnings',
    graph=gml_string
)


# Get the estimand
estimand = model.identify_effect()

# Get estimate (Linear Regression)
estimate = model.estimate_effect(
    identified_estimand=estimand,
    method_name="backdoor.distance_matching",
    target_units="ate",
    method_params={'distance_metric': "minkowski", 'p':2})

Expected behavior It should not matter what encoding we use for binary treatment or it should be explicitly specified that only bool-typed values are accepted and error message should reflect that.

Version information:

• DoWhy version 0.8

Additional context