StatsModelsLinearRegression needs check_input

[moprescu]

If the numpy arrays passed are of type object and not float, then an exception is thrown in param, _, rank, _ = np.linalg.lstsq(WX, wy, rcond=None)

[kbattocchi]

Simple reproduction:

from econml.sklearn_extensions.linear_model import StatsModelsLinearRegression
from sklearn.linear_model import LinearRegression
import numpy as np

X = np.random.normal(size=(100,5)).astype('O')
y = np.random.normal(size=100)

LinearRegression().fit(X,y)
StatsModelsLinearRegression().fit(X,y)