Closed wilsonrljr closed 3 years ago
Currently, to build NAR or NARMA models the user must create a input with zeros and pass it to the fit method.
Alternative
Instead of using
x_train = np.zeros_like(y_train) model.fit(x_train, y_train)
the fit method for NAR models should be
x_train = np.zeros_like(y_train) model.fit(y_train)
Currently, to build NAR or NARMA models the user must create a input with zeros and pass it to the fit method.
Alternative
Instead of using
the fit method for NAR models should be