49
93%

Ran 14 Jan 2014 10:52PM UTC

Jobs 1

Files 128

Run time –

Badge

Embed ▾

pending completion

Build # 49

Build Type

push

travis-ci

Committed by

matthew-brett

Commit Message

Merge pull request #211 from matthew-brett/add-proxy-slicing

MRG: add proxy slicing

Long awaited API to be able to do arbitrary slicing on array proxies from images.

Meaning, you can slice data from images without loading the whole image from disk.

Of course this can save memory, but it can also be much faster. For an
uncompressed nifti image:

In [17]: timeit img = nib.load(fname1); arr = img.dataobj[..., 120]
1000 loops, best of 3: 1.7 ms per loop

In [19]: timeit img = nib.load(fname1); arr = img.get_data()[..., 120]
1 loops, best of 3: 713 ms per loop
For a gzipped image, the savings depend on the position in the file:

In [8]: timeit img = nib.load(fname); arr = img.dataobj[..., 2]
10 loops, best of 3: 30.7 ms per loop

In [7]: timeit img = nib.load(fname); arr = img.get_data()[..., 2]
1 loops, best of 3: 1.32 s per loop

In [12]: timeit img = nib.load(fname); arr = img.dataobj[..., 40]
1 loops, best of 3: 466 ms per loop

In [11]: timeit img = nib.load(fname); arr = img.get_data()[..., 40]
1 loops, best of 3: 1.31 s per loop

In [13]: timeit img = nib.load(fname); arr = img.dataobj[..., 100]
1 loops, best of 3: 1.15 s per loop

In [14]: timeit img = nib.load(fname); arr = img.dataobj[..., 120]
1 loops, best of 3: 1.38 s per loop

Run Details

13990 of 15469 relevant lines covered (90.44%)

0.9 hits per line

Jobs

ID	Job ID	Ran	Files	Coverage
4	49.4 (COVERAGE=--with-coverage)	14 Jan 2014 10:52PM UTC	0	90.44	Travis Job 49.4

Source Files on build 49

Detailed source file information is not available for this build.

nipy / nibabel / 49
93%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 49

nipy / nibabel / 49 93%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 49

nipy / nibabel / 49
93%

README BADGES
x