Mailman 3 July 2018 - scikit-image

Job Posting-Scientific Software Developer Position at Enthought
by Jill Cowan July 31, 2018

July 31, 2018

Enthought is currently recruiting for a scientific software developer with experience in image processing. If you might be interested, please see https://www.enthought.com/careers/. If you have any questions, feel free to contact Jill Cowan at jobs(a)enthought.com. _______________ Jill Cowan Operations and Recruiting | Enthought Tel +1 512 536 1057 jillc(a)enthought.com <jillc(a)enthought.com>

1 0

Re: [scikit-image] Request for comment: road to scikit-image 1.0
by Michael Aye July 30, 2018

July 30, 2018

Just in terms of descriptors I always find “tool set” to give the impression of being immensely useful, one could garnish it with “essential”. ;) Sent from my iPhone > On Jul 30, 2018, at 09:43, Juan Nunez-Iglesias <jni(a)fastmail.com> wrote: > > Hi Everyone, > > A couple of weeks ago I published this blog post: > > https://ilovesymposia.com/2018/07/13/the-road-to-scikit-image-1-0/ > > I didn't publicise it widely because it was in the middle of SciPy 2018 and I had many other things pulling at my attention. I really want to drive this vision forward, but not without further input from the community (ie you guys). So I would really appreciate if you could take a few minutes to read it, then publish your criticisms/comments/wishlist to: > > https://github.com/scikit-image/scikit-image/issues/3263 > > You can also just respond to this list. > > I'll prime the discussion with a criticism from Stéfan: "reference" library is not an exciting word to describe scikit-image. What 1-2 adjectives would you use to describe your vision for scikit-image? We want to convey readability, ease of use, accuracy, and (reasonable) performance. > > After some reasonable number of comments, I'll do my best to compile them and submit a PR to our documentation that will enshrine the document within an aura of authority. =P > > Thanks! > > Juan. > > PS: The whole discussion is rather timely: Egor Panfilov has drawn our attention (on gitter) to the fact that we have just passed our 10,000th commit! > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image

1 0

Request for comment: road to scikit-image 1.0
by Juan Nunez-Iglesias July 30, 2018

July 30, 2018

Hi Everyone, A couple of weeks ago I published this blog post: https://ilovesymposia.com/2018/07/13/the-road-to-scikit-image-1-0/ I didn't publicise it widely because it was in the middle of SciPy 2018 and I had many other things pulling at my attention. I really want to drive this vision forward, but not without further input from the community (ie you guys). So I would really appreciate if you could take a few minutes to read it, then publish your criticisms/comments/wishlist to: https://github.com/scikit-image/scikit-image/issues/3263 You can also just respond to this list. I'll prime the discussion with a criticism from Stéfan: "reference" library is not an exciting word to describe scikit-image. What 1-2 adjectives would you use to describe your vision for scikit-image? We want to convey readability, ease of use, accuracy, and (reasonable) performance. After some reasonable number of comments, I'll do my best to compile them and submit a PR to our documentation that will enshrine the document within an aura of authority. =P Thanks! Juan. PS: The whole discussion is rather timely: Egor Panfilov has drawn our attention (on gitter) to the fact that we have just passed our 10,000th commit!

1 0

Re: [scikit-image] Installing from pypi tarball without cython
by Stefan van der Walt July 29, 2018

July 29, 2018

I don't feel strongly about it, we just need to update our statement about what is required in the documentation. And the pending PR should help for pip, at least. Best regards, Stéfan

2 1

Re: [scikit-image] Installing from pypi tarball without cython
by Stefan van der Walt July 28, 2018

July 28, 2018

On Fri, 27 Jul 2018 13:31:36 -0400, Matthew McCormick wrote: > I created a PR to address this issue here: > > https://github.com/scikit-image/scikit-image/pull/3295 Thanks, Matt, this is useful. That said, I think Stuart's bug report is still valid: the source tarball is supposed to ship with precompiled Cython code, and thus should not require Cython to be present in order to build. Stéfan

2 1

Re: [scikit-image] [JOB] Microscopy Image Analyst position at Fred Hutch (Seattle)
by Stefan van der Walt July 28, 2018

July 28, 2018

On Fri, 27 Jul 2018 20:06:41 +0300, Egor Panfilov wrote: > I wonder what is this mailing list policy on job postings? Probably, there > are many interesting vacancies whirling around our community, and some of > them are more than just relevant. I'd be also interesting in sharing some, > occasionally (potentially, in-academia ones). I think any image processing-related job that explicitly mentions either Python or scikit-image is on topic. Any company that sponsors scikit-image developer time for their employees can get space on the website to advertise! Best regards, Stéfan

1 0

Re: [scikit-image] [JOB] Microscopy Image Analyst position at Fred Hutch (Seattle)
by Juan Nunez-Iglesias July 28, 2018

July 28, 2018

My personal preference is to allow well-targeted job postings (as this one) on the list. > On 28 Jul 2018, at 3:06 am, Egor Panfilov <egor.v.panfilov(a)gmail.com> wrote: > > Hi Ariel, > Thank you for posting! > > Stefan, > I wonder what is this mailing list policy on job postings? Probably, there are many interesting vacancies whirling around our community, and some of them are more than just relevant. I'd be also interesting in sharing some, occasionally (potentially, in-academia ones). > In order to bring some food to the discussion, very large projects (such as OCaml, for example) typically have a dedicated `something`-jobs mailing lists. The current bandwidth of this list, though, is not too high, so it doesn't seem to me to have sense to adapt that approach. > > Regards, > Egor > > On Fri, 27 Jul 2018 at 19:27, Ariel Rokem <arokem(a)gmail.com <mailto:arokem@gmail.com>> wrote: > With apologies for cross-posting, this position at the Fred Hutch Cancer Research Center in Seattle may be of interest to some on this list: > > https://careers-fhcrc.icims.com/jobs/11530/morphometry-and-image-analyst/jo… <https://careers-fhcrc.icims.com/jobs/11530/morphometry-and-image-analyst/jo…> > > If you have any questions regarding this position, please contact Emily Hatch (ehatch(a)fredhutch.org <mailto:ehatch@fredhutch.org>). > > Cheers, > > Ariel > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org <mailto:scikit-image@python.org> > https://mail.python.org/mailman/listinfo/scikit-image <https://mail.python.org/mailman/listinfo/scikit-image> > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image

1 0

Re: [scikit-image] Installing from pypi tarball without cython
by Stuart Mumford July 27, 2018

July 27, 2018

Hi Egor, Thanks for the response. I have no issues with compiling the C code, I have a full compile chain setup on the machine etc. What I was expecting to happen is that the C code generated by Cython would be included in the tarball and therefore not need Cython to build from a tarball. This assumption is mainly based on the error that was printed: "RuntimeError: Cython >= 0.23 is required to build scikit-image from git checkout" which indicates (and the fact it's not in the package requirements) Cython should not be needed to build from PyPI. I hope that makes it a little clearer, Thanks again, Stuart On Friday, 27 July 2018 12:02:46 BST Egor Panfilov wrote: > Hi Stuart, > > I don't think what you are experiencing is an issue. Several components > (i.e. functions) of `scikit-image` are implemented in Cython, and require > it in order to be build and linked with Python API. "Binary" distributions > of `scikit-image` (including the PyPi .whl-files) ship such components > pre-built, so the requirements for running `scikit-image` and building it > are different. You can take a look at the requirements for different > usecases (building package, executing package, building gallery, etc) in > our repository in > https://github.com/scikit-image/scikit-image/tree/master/requirements . > If you are only interested in installing `scikit-image`, I'd you recommend > you to use the provided pre-built distributions available for the most > popular architectures and operating systems: > pip install --no-cache-dir scikit-image > > Regards, > Egor > > On Fri, 27 Jul 2018 at 13:28, Stuart Mumford <stuart(a)cadair.com> wrote: > > Hello, > > > > > > > > I have just run into an error while trying to install scikit-image from > > the pip tarball on linux (as part of the SunPy test infrastructure). It > > seems that the PyPI tarball needs Cython to be installed: > > > > > > > >  pip install --no-binary=:all: --no-cache-dir scikit-image > > Collecting scikit-image > > > > Downloading > > > > https://files.pythonhosted.org/packages/fc/20/d3e736493b16e9455ce8579722d6 > > 44b313814c599d5824d34e448845f746/scikit-image-0.14.0.tar.gz (27.0MB) > > > > 100% |████████████████████████████████| 27.0MB 5.9MB/s > > Complete output from command python setup.py egg_info: > > Partial import of skimage during the build process. > > > > Traceback (most recent call last): > > File "/tmp/pip-install-ypyllick/scikit-image/skimage/_build.py", line > > > > 30, in cython > > > > from Cython import __version__ > > > > ModuleNotFoundError: No module named 'Cython' > > > > During handling of the above exception, another exception occurred: > > > > Traceback (most recent call last): > > File "<string>", line 1, in <module> > > File "/tmp/pip-install-ypyllick/scikit-image/setup.py", line 149, in > > > > <module> > > > > **extra > > > > File > > > > "/opt/miniconda/envs/clean/lib/python3.7/site-packages/numpy/distutils/cor > > e.py", line 135, in setup > > > > config = configuration() > > > > File "/tmp/pip-install-ypyllick/scikit-image/setup.py", line 70, in > > > > configuration > > > > config.add_subpackage('skimage') > > > > File > > > > "/opt/miniconda/envs/clean/lib/python3.7/site-packages/numpy/distutils/mis > > c_util.py", line 1034, in add_subpackage > > > > caller_level = 2) > > > > File > > > > "/opt/miniconda/envs/clean/lib/python3.7/site-packages/numpy/distutils/mis > > c_util.py", line 1003, in get_subpackage > > > > caller_level = caller_level + 1) > > > > File > > > > "/opt/miniconda/envs/clean/lib/python3.7/site-packages/numpy/distutils/mis > > c_util.py", line 940, in _get_configuration_from_setup_py > > > > config = setup_module.configuration(*args) > > > > File "skimage/setup.py", line 14, in configuration > > > > config.add_subpackage('feature') > > > > File > > > > "/opt/miniconda/envs/clean/lib/python3.7/site-packages/numpy/distutils/mis > > c_util.py", line 1034, in add_subpackage > > > > caller_level = 2) > > > > File > > > > "/opt/miniconda/envs/clean/lib/python3.7/site-packages/numpy/distutils/mis > > c_util.py", line 1003, in get_subpackage > > > > caller_level = caller_level + 1) > > > > File > > > > "/opt/miniconda/envs/clean/lib/python3.7/site-packages/numpy/distutils/mis > > c_util.py", line 940, in _get_configuration_from_setup_py > > > > config = setup_module.configuration(*args) > > > > File "skimage/feature/setup.py", line 22, in configuration > > > > cython(['_haar.pyx'], working_path=base_path) > > > > File "/tmp/pip-install-ypyllick/scikit-image/skimage/_build.py", line > > > > 42, in cython > > > > % CYTHON_VERSION) > > > > RuntimeError: Cython >= 0.23 is required to build scikit-image from git > > > > checkout > > > > Cython >= 0.23 not found; falling back to pre-built geometry.c > > Cython >= 0.23 not found; falling back to pre-built transform.c > > Cython >= 0.23 not found; falling back to pre-built interpolation.c > > Cython >= 0.23 not found; falling back to pre-built _draw.c > > Cython >= 0.23 not found; falling back to pre-built corner_cy.c > > Cython >= 0.23 not found; falling back to pre-built censure_cy.c > > Cython >= 0.23 not found; falling back to pre-built orb_cy.c > > Cython >= 0.23 not found; falling back to pre-built brief_cy.c > > Cython >= 0.23 not found; falling back to pre-built _texture.c > > Cython >= 0.23 not found; falling back to pre-built _hessian_det_appx.c > > Cython >= 0.23 not found; falling back to pre-built _hoghistogram.c > > > > ---------------------------------------- > > > > Command "python setup.py egg_info" failed with error code 1 in > > /tmp/pip-install-ypyllick/scikit-image/ > > > > I don't know if this is a bug in the tarball / setup machinery or if it's > > some silly mistake I am making. > > > > > > > > Thanks for any help, > > > > Stuart > > _______________________________________________ > > scikit-image mailing list > > scikit-image(a)python.org > > https://mail.python.org/mailman/listinfo/scikit-image

2 1

Re: [scikit-image] [JOB] Microscopy Image Analyst position at Fred Hutch (Seattle)
by Egor Panfilov July 27, 2018

July 27, 2018

Hi Ariel, Thank you for posting! Stefan, I wonder what is this mailing list policy on job postings? Probably, there are many interesting vacancies whirling around our community, and some of them are more than just relevant. I'd be also interesting in sharing some, occasionally (potentially, in-academia ones). In order to bring some food to the discussion, very large projects (such as OCaml, for example) typically have a dedicated `something`-jobs mailing lists. The current bandwidth of this list, though, is not too high, so it doesn't seem to me to have sense to adapt that approach. Regards, Egor On Fri, 27 Jul 2018 at 19:27, Ariel Rokem <arokem(a)gmail.com> wrote: > With apologies for cross-posting, this position at the Fred Hutch Cancer > Research Center in Seattle may be of interest to some on this list: > > > https://careers-fhcrc.icims.com/jobs/11530/morphometry-and-image-analyst/jo… > > If you have any questions regarding this position, please contact Emily > Hatch (ehatch(a)fredhutch.org). > > Cheers, > > Ariel > _______________________________________________ > scikit-image mailing list > scikit-image(a)python.org > https://mail.python.org/mailman/listinfo/scikit-image >

1 0

[JOB] Microscopy Image Analyst position at Fred Hutch (Seattle)
by Ariel Rokem July 27, 2018

July 27, 2018

With apologies for cross-posting, this position at the Fred Hutch Cancer Research Center in Seattle may be of interest to some on this list: https://careers-fhcrc.icims.com/jobs/11530/morphometry-and-image-analyst/jo… If you have any questions regarding this position, please contact Emily Hatch (ehatch(a)fredhutch.org). Cheers, Ariel

1 0