How to mix onehotencoding and bagofword

Question

I am looking for a model, a sort of onehotencoding, that could mix columns of the same category in a single vector.

• Having the Data [5,8,1,3]
• it will gives me : [0,1,0,1,0,1,0,0,1,0,0,0]

with an arbitrary size set at 12.

I looked at bagofword but I did not find how to set the vector size independently of the input data.

If somebody can gives me some clues, it will be find.

Answer 1

Note that Bag-of-words models are used when dealing with text. For this simpler task you can just use np.bincount and specify a minlength:

l = [5,8,1,3]

np.bincount(l, minlength=12)
# array([0., 1., 0., 1., 0., 1., 0., 0., 1., 0., 0., 0.])