CODEWITHSUNDEEP
pythonmachine-learningscikit-learncross-validationxgboost
Isac Moura
Isac Moura

Reputation: 6898

How to pass X_train + X_eval to Random Search CV fit method

I'm trying to pass a set of [X_train, X_val] to X of random_search.fit() and the same to y with:
random_search.fit(X=[X_train, X_val], y=[y_train, y_val])

But when the training occurs the below error is displayed:

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()
During handling of the above exception, another exception occurred:
TypeError: can not initialize DMatrix from list

I already tried using group field of fit method but I got another error. There any way of pass train/test and his evaluations sets for the fit method of Random Search? I can't figure out how to do this.

Upvotes: 0

Views: 290

Answers (1)

Raphaele Adjerad
Raphaele Adjerad

Reputation: 1125

Maybe I have not understood correctly but if you want to use RandomizedSearchCV on the training and the validation samples concatenated together and use CV on this whole sample, I suggest using np.concatenate instead of list comprehension as follows:

# taking examples for your X_train, X_val, y_train and y_val
import numpy as np
X_train = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
X_val = np.array([[1, 2, 3], [4, 5, 6]])
y_train = np.array([10, 11, 12])
y_val = np.array([13, 14])

data = np.concatenate((X_train, X_val), axis=0)
target = np.concatenate((y_train, y_val))

And you can pass data and target to the fit method.

Upvotes: 2

Related Questions