create ndarray from numpy memmap slice

Question

I am wondering how to determine if an object is a direct memmap object or a slice descendent of one.

The easiest way to as the question is with an example:

>>> import numpy as np
>>> filename = '../sandbox/test.bin'
>>> a = np.memmap(filename, dtype='float32', offset=0, shape=(4,2), order='F')
>>> print a
[[ 1.  5.]
 [ 2.  6.]
 [ 3.  7.]
 [ 4.  8.]]
>>> a.filename
'Z:\\CNI\\sandbox\\test.bin'
>>> a.shape
(4L, 2L)
>>> a.offset
0
>>>
>>> b = a[:,1]
>>> print b
[ 5.  6.  7.  8.]
>>> b.filename
'Z:\\CNI\\sandbox\\test.bin'
>>> b.shape
(4L,)
>>> b.offset
0
>>>

how would I determine that b is not a memmap object, but is instead a slice descendent of one? Or at the very least that b's offset is wrong. (in this example the offset should be 4)

Check out the function get_data_base from here. I'm not 1000% sure it works with memmaps, but a priori it should. — eickenberg
– eickenberg, Commented Nov 13, 2014 at 9:50
@user3701099 in this example you can check the OWNDATA parameter of the b.flags attribute, which is False for a slice... — Saullo G. P. Castro
– Saullo G. P. Castro, Commented Nov 13, 2014 at 14:31
@SaulloCastro the OWNDATA in flags attribute works and is very straightforward — Joshua
– Joshua, Commented Nov 13, 2014 at 17:39
@eickenberg the get_data_base function also works from the link you post — Joshua
– Joshua, Commented Nov 13, 2014 at 17:40

Saullo G. P. Castro · Accepted Answer · 2014-11-13 19:08:49Z

1

In this example you can check the OWNDATA parameter of the:

b.flags

attribute, which is False for a slice...

answered Nov 13, 2014 at 19:08

Saullo G. P. Castro

59.4k28 gold badges191 silver badges244 bronze badges

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

Collectives™ on Stack Overflow

create ndarray from numpy memmap slice

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