python: list of matrices to numpy array?

Question

I've got a list containing numpy matrices. Anyway that I could turn the whole thing into a nice clean numpy array?

From:

[matrix([[1]]), matrix([[ 1.99387871]]), matrix([[ 2.53564618]]), matrix([[ 4.39125807]]), matrix([[ 4.246309]]), matrix([[ 5.21571607]]), matrix([[ 6.17408811]]), matrix([[ 4.75146571]]), matrix([[ 6.19319742]]), matrix([[ 6.1277607]]), matrix([[ 7.43821216]])]

To:

[[1 1.99387871 2.53564618 4.39125807 4.246309 5.21571607 6.17408811 4.75146571 6.19319742 6.1277607 7.43821216]]

M4rtini · Accepted Answer · 2013-12-23 19:35:54Z

6

b = np.asarray(a, dtype=float)
#to get the same shape do.
b = b.reshape(-1, len(b)) 
#to just get one dimmension do. 
b = np.asarray(a, dtype=float).reshape(len(a))

edited Dec 23, 2013 at 19:35

answered Dec 23, 2013 at 19:24

M4rtini

13.6k4 gold badges37 silver badges42 bronze badges

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

3 Comments

DSM Over a year ago

Maybe add .reshape(-1, len(a)) if the OP cares about the shape.

abarnert Over a year ago

Also, I don't think you want the dtype=float. Without it, you preserve the dtype of the individual matrices—or, if they're different, as in the OP's example, the coerced union of the types. That's probably better than forcing float even if there were, say, float128 or complex128 values in the matrices.

M4rtini Over a year ago

@abarnert True, for the general case forcing float is not a good solution, but setting the dtype when you know what it is, usually a good idea. Sometimes numpy isn't smart enough to figure them out and will just set it as a generic object. which can cause problems.

Collectives™ on Stack Overflow

python: list of matrices to numpy array?

1 Answer 1

3 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related