Mailman 3 October 2013 - scikit-image

Re: Release time
by Stéfan van der Walt 19 Oct '13

19 Oct '13

On Sat, Oct 19, 2013 at 12:37 PM, Ankit Agrawal <aaaagrawal(a)gmail.com> wrote: > Any updates about the 0.9 release? Just waiting for GitHub to come back online to push the tags. Stéfan

1 0

Re: Release time
by Ankit Agrawal 19 Oct '13

19 Oct '13

Any updates about the 0.9 release? Regards, Ankit Agrawal, Communication and Signal Processing, IIT Bombay. On Thu, Oct 17, 2013 at 10:27 AM, Juan Nunez-Iglesias <jni.soma(a)gmail.com>wrote: > Hi Stefan, > > It just occurred to me that this comment of mine: > https://github.com/scikit-image/scikit-image/pull/674#issuecomment-22308750 > > was ignored. As I mention in the comment, the refactoring of regionprops > actually resulted in an API change. Seems fixable... Do you want me to > tackle it? Here's the error it gives: > > /Users/nuneziglesiasj/venv/husc/lib/python2.7/site-packages/husc/features.py in object_features(bin_im, im, erode) > 141 'MaxIntensity', 'Solidity'], > 142 intensity_image=im) > --> 143 feats = np.array([props.values() for props in feats], np.float) > 144 feature_quantiles = mquantiles(feats, [0.05, 0.25, 0.5, 0.75, 0.95], > 145 axis=0) > > AttributeError: '_RegionProperties' object has no attribute 'values' > > > I think it would be fixable by adding a .values() function (which would be > deprecated from the beginning) that returns whatever regionprops were > originally requested. > > @ahojnnes might have something to add... > > Juan. > > > > On Thu, Oct 17, 2013 at 12:11 AM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote: > >> On Tue, Oct 15, 2013 at 3:25 PM, Johannes Schönberger <jsch(a)demuc.de> >> wrote: >> > I think I fixed the two remaining issues reported by Christoph, thanks >> for reporting them! >> >> Alright! Everything's gone quiet... so I'm starting the countdown to >> the release. Speak now or hold your deprecations until 0.10! >> >> Stéfan >> >> -- >> You received this message because you are subscribed to the Google Groups >> "scikit-image" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to scikit-image+unsubscribe(a)googlegroups.com. >> For more options, visit https://groups.google.com/groups/opt_out. >> > > -- > You received this message because you are subscribed to the Google Groups > "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. >

1 0

`min_size` argument in Feltzenswalb segmentation has no effect
by Guillaume Gay 17 Oct '13

17 Oct '13

Hi all, The title says it, with a recent 0.9dev version of skimage, I can't see an effect of the `min_size` argument in Felzenszwalb segmentation. BTW the meaning of this argument is not very clear in the doc. It's the first time I use this function, so I don't know wether this is a regression or not. Reading the code, I fail to understand how this all works.... Here is a notebook showing the (lack of) effect of min_size: http://nbviewer.ipython.org/7023716 Cheers, good luck with the realease Guillaume

2 1

Re: Release time
by Juan Nunez-Iglesias 17 Oct '13

17 Oct '13

Hi Stefan, It just occurred to me that this comment of mine: https://github.com/scikit-image/scikit-image/pull/674#issuecomment-22308750 was ignored. As I mention in the comment, the refactoring of regionprops actually resulted in an API change. Seems fixable... Do you want me to tackle it? Here's the error it gives: /Users/nuneziglesiasj/venv/husc/lib/python2.7/site-packages/husc/features.py in object_features(bin_im, im, erode) 141 'MaxIntensity', 'Solidity'], 142 intensity_image=im) --> 143 feats = np.array([props.values() for props in feats], np.float) 144 feature_quantiles = mquantiles(feats, [0.05, 0.25, 0.5, 0.75, 0.95], 145 axis=0) AttributeError: '_RegionProperties' object has no attribute 'values' I think it would be fixable by adding a .values() function (which would be deprecated from the beginning) that returns whatever regionprops were originally requested. @ahojnnes might have something to add... Juan. On Thu, Oct 17, 2013 at 12:11 AM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote: > On Tue, Oct 15, 2013 at 3:25 PM, Johannes Schönberger <jsch(a)demuc.de> > wrote: > > I think I fixed the two remaining issues reported by Christoph, thanks > for reporting them! > > Alright! Everything's gone quiet... so I'm starting the countdown to > the release. Speak now or hold your deprecations until 0.10! > > Stéfan > > -- > You received this message because you are subscribed to the Google Groups > "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. >

1 0

Re: Release time
by Johannes Schönberger 16 Oct '13

16 Oct '13

I think I fixed the two remaining issues reported by Christoph, thanks for reporting them! > How about we make a 0.9.x branch where we cherry-pick any serious bugs > affecting 0.9.0? Good idea.

2 1

Operations on disk-based images
by Francesc Alted 15 Oct '13

15 Oct '13

Hi, We are trying to deal with very large images (they typically do not fit in-memory), and for visualizing purposes, I am trying to use some downsampling before sending the bits to the graphics render. Our files are typically in HDF5 (or NetCDF4) format, so the data on-disk is accessible randomly (i.e. they support indexes a la NumPy), and I noticed that the recently added `skimage.transform.downscale_local_mean()` function (which is just perfect for our purposes) also works with these on-disk arrays. The problem is that the function takes *ages* to finish. My guess is that this is mainly a consequence of the fact that accessing data on-disk without too much care about data locality, is much more expensive than doing the same in-memory. I suppose the only solution is to rewrite the algorithms we are interested in so that leveraging spatial locality would be critical, but I wanted to report that here just in case someone has some insight on what to do in this case. Thanks, -- Francesc Alted

1 0

Re: Release time
by Marianne Corvellec 15 Oct '13

15 Oct '13

Hi anyone-with-time-this-weekend, Please let me know whether my last commit wraps it up for that still open issue: https://github.com/scikit-image/scikit-image/pull/624 Somehow I don't get the expected warnings when calling deprecated functions... x-( Happy Canadian Thanksgiving! Marianne On Friday, October 11, 2013 12:40:14 PM UTC-4, Stefan van der Walt wrote: > > On Fri, Oct 11, 2013 at 5:13 PM, Johannes Schönberger <js...(a)demuc.de<javascript:>> > wrote: > > Do we plan to push 0.9 to the public this weekend? > > I finished the Debian packaging last weekend, so there's nothing > stopping us from a distribution point of view. > > > There are only a few minor outstanding issues: > https://github.com/scikit-image/scikit-image/issues?milestone=2&state=open > > I'll work through these... everyone who has got some spare cycles, > please glance at it! > > Thanks > Stéfan >

5 7

Re: Build error
by Johannes Schönberger 14 Oct '13

14 Oct '13

Try ``make clean``. Am 14.10.2013 um 18:06 schrieb Jean K <jean.kossaifi(a)gmail.com>: > Yes > % cython -V > Cython version 0.19.2 > > I did python setup.py clean before installing but still get the same error... > > > 2013/10/14 Stéfan van der Walt <stefan(a)sun.ac.za> > On Mon, Oct 14, 2013 at 5:21 PM, Jean K <jean.kossaifi(a)gmail.com> wrote: > > You are absolutely right, thanks! > > I had already installed Cython locally but didn't add it properly to the > > PATH. > > > > However after I fixed it I got another error: > > Do you have Cython v0.19 installed? If so, then perhaps try a clean build. > > Stéfan > > -- > You received this message because you are subscribed to a topic in the Google Groups "scikit-image" group. > To unsubscribe from this topic, visit https://groups.google.com/d/topic/scikit-image/zV05Huop18c/unsubscribe. > To unsubscribe from this group and all of its topics, send an email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > > -- > You received this message because you are subscribed to the Google Groups "scikit-image" group. > To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out.

2 1

Re: Build error
by Jean K 14 Oct '13

14 Oct '13

Yes > % cython -V > Cython version 0.19.2 I did python setup.py clean before installing but still get the same error... 2013/10/14 Stéfan van der Walt <stefan(a)sun.ac.za> > On Mon, Oct 14, 2013 at 5:21 PM, Jean K <jean.kossaifi(a)gmail.com> wrote: > > You are absolutely right, thanks! > > I had already installed Cython locally but didn't add it properly to the > > PATH. > > > > However after I fixed it I got another error: > > Do you have Cython v0.19 installed? If so, then perhaps try a clean build. > > Stéfan > > -- > You received this message because you are subscribed to a topic in the > Google Groups "scikit-image" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/scikit-image/zV05Huop18c/unsubscribe. > To unsubscribe from this group and all of its topics, send an email to > scikit-image+unsubscribe(a)googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. >

1 0

Richardson-Lucy deconvolution
by Christoph Deil 14 Oct '13

14 Oct '13

Hi, I wanted to try out Richardson-Lucy deconvolution on some astronomy images. Is this already available in some scientific Python package? Would it be a welcome addition for scikit-image? It seems very easy to implement looking at this description and I could give it a try. http://en.wikipedia.org/wiki/Richardson–Lucy_deconvolution Christoph

4 3