Convert the following code from Matlab to Python:

all_dist = pdist2(test_image_feats, train_image_feats, 'euclidean');

[sorted_dist, index] = sort(all_dist, 2);

predicted_categories = train_labels(index(:, 1),:);

test_image_feats and train_image_feats are M x d matrices.

train_labels is an N x 1 cell array, where each entry is a string

I've converted the first line to:

D = sklearn_pairwise.pairwise_distances(test_image_feats, train_image_feats)

But I do not know how to do the last two lines.