How to concatenate features of different types in a single python array?

Question

I have a features array that contains values of different types:

>>> features = train_df.values
>>> [x for x in features]

[True,
 array([2, 0, 0, ..., 0, 0, 0]),
 False,
 False,
 17,
 1,
 10,
 array([0, 0, 0, ..., 0, 0, 0])]

I would like to produce a single python array that contains a concatenation of the all of the above features, i.e.

np.array([True, 2, 0, 0, ..., 0, 0, 0, False, False, 17, 1, 10, 0, 0, 0, ..., 0, 0, 0])

My goal is to train sklearn LogisticRegression with the above feature vector. What is the best way to do this in python?

cs95 · Accepted Answer · 2017-10-22 02:48:41Z

3

You could do this with a simple list comprehension.

>>> x
[True, array([2, 0, 0, 0, 0, 0]), False, False, 17, 1, 10, array([0, 0, 0, 0, 0, 0])]

>>> [j for i in x for j in (i if isinstance(i, np.ndarray) else (i, ))]
[True, 2, 0, 0, 0, 0, 0, False, False, 17, 1, 10, 0, 0, 0, 0, 0, 0]

>>> np.array(_, dtype='O')
array([True, 2, 0, 0, 0, 0, 0, False, False, 17, 1, 10, 0, 0, 0, 0, 0, 0], dtype=object)

If you don't add dtype='O', your bools will be casted to integers. It's upto you whether you want that or not. Working with object arrays are usually frowned upon, since they provide no vectorisation/efficiency benifits.

edited Oct 22, 2017 at 2:48

answered Oct 22, 2017 at 2:40

cs95

406k106 gold badges744 silver badges797 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

How to concatenate features of different types in a single python array?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related