[FIXED] Sklearn inverse transform LabelBinarizer for only one encoded value

Issue

I have successfully encoded my data like this:

reviewerLe = LabelBinarizer()
reviewerIDs = reviewerLe.fit_transform(df['reviewerID'])

and

print(reviewerIDs[1])
reviewerIDs[1].shape

outputs

[0 0 0 ... 0 0 0]
(4975,)

I am wondering how to inverse transform only one or a couple of predicted values.

reviewerLe.inverse_transform(reviewerIDs[1])

outputs

axis 1 is out of bounds for array of dimension 1

I don't quite understand what the function exprects. In the docs is stated :

inverse_transform(Y, threshold=None) Y{ndarray, sparse matrix} of shape (n_samples, n_classes)

Target values. All sparse matrices are converted to CSR before inverse transformation.

Does that mean that i should reshape my array to (1,4975)? If yes, how?

Solution

inverse_transform() expects a 2D array, but you're passing a 1D array to it.

You can reshape your 1D array into a 2D

reviewerIDs_2D = reviewerIDs[1].reshape(1, -1)
reviewerLe.inverse_transform(reviewerIDs_2D)

Answered By - Alperen Tahta