Suspected coding mistake in PerformanceEvaluator

[prabaey]

Description

There seems to be a coding mistake in the function _evaluate_performance_regression within the class PerformanceEvaluator. I think I spotted a typo in the transformation of the training and test set to numpy arrays, which leads to the estimator being trained and tested on the same data.

How to Reproduce

Simply investigate the source code for the function _evaluate_performance_regression within the class PerformanceEvaluator in the file eval_performance.py.

Expected Behavior

The following snippet of code

 X_train = np.asarray(X_test)
 X_test = np.asarray(X_test)
 y_train = np.asarray(y_test)
 y_test = np.asarray(y_test)

should be changed to:

 X_train = np.asarray(X_train)
 X_test = np.asarray(X_test)
 y_train = np.asarray(y_train)
 y_test = np.asarray(y_test)

[robsdavis]

@prabaey, thanks for posting this issue! It does indeed seem to be a mistake. A PR to fix this will be created shortly.