abria / TeraStitcher

A tool for fast automatic 3D-stitching of teravoxel-sized microscopy images
http://abria.github.io/TeraStitcher/
Other
78 stars 32 forks source link

Updated Parastitcher3.2.0.pyc for Python 3.7 #52

Open microscopy-jacksonlabs opened 4 years ago

microscopy-jacksonlabs commented 4 years ago

Hi, I am trying the mpiexec python script for Windows and are wondering if you would be able to provide an updated parastitcher pyc file to be compatible with python3.7 and 3.7.1. Thank you for the great work on TeraStitcher.

iannellog commented 4 years ago

Attached there is a .zip containing the last version of the script compiled with Python 3.4. Let me know if it works. paraconverter2_3_2.pyc.zip

microscopy-jacksonlabs commented 4 years ago

Thank you Giulio for your quick response. I tried this using python 3.4.0, which allows the use of pip and getting mpi4py, but this results in the same magic number error.

Unfortunately, pip install does not work (anymore) with 3.2.0 or 3.2.5, as no ‘pip<8’ is available. I tried hex-editing the magic numbers for 3.4 or 3.2 in boht pyc’s, but that does not work for the mpi support,a s it results in another error. Last, I will try to manually install mpi4py for 3.2.5, but that is a bit elaborate and may better work on linux.

Just as background, my organization tries to use TS for very large 3D tiles from an multi-photon instrument (and light sheet, a tlower resolution) and I would like to work on an automated pipeline for screening several hundreds of tissues. The mpiexec/mpirun seems to be a feasible process to add on superb cuda support since 1.11, but of course this little glitch is a current setback.

Thank you for your help.

Philipp

Philipp P. Henrich, PhD Senior Microscopy Scientist The Jackson Laboratory B1 1250 600 Main Street Bar Harbor, Maine 04609

207.288.6000,1702 t philipp.henrich@jax.orgmailto:philipp.henrich@jax.org

The Jackson Laboratory Bar Harbor, ME | Farmington, CT | Sacramento, CA www.jax.orghttp://www.jax.org/

From: Giulio Iannello [mailto:notifications@github.com] Sent: Wednesday, September 04, 2019 5:28 AM To: abria/TeraStitcher Cc: Philipp Henrich; Author Subject: Re: [abria/TeraStitcher] Updated Parastitcher3.2.0.pyc for Python 3.7 (#52)

Attached there is a .zip containing the last version of the script compiled with Python 3.4. Let me know if it works. paraconverter2_3_2.pyc.ziphttps://github.com/abria/TeraStitcher/files/3573772/paraconverter2_3_2.pyc.zip

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/abria/TeraStitcher/issues/52?email_source=notifications&email_token=ANCRT4ZPBTWV5APMWTIIHXDQH55RPA5CNFSM4ITJCYD2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD526TNI#issuecomment-527821237, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ANCRT47KIH3TH5XW5YB37OTQH55RPANCNFSM4ITJCYDQ.

The information in this email, including attachments, may be confidential and is intended solely for the addressee(s). If you believe you received this email by mistake, please notify the sender by return email as soon as possible.

chrisroat commented 4 years ago

Are the parastitcher/paraconverter scripts able to be checked into the repo? If so, perhaps we can avoid the zip-file downloads and the community could commit fixes to have them work in modern python versions.

chrisroat commented 4 years ago

Please consider updating the para* scripts for python3.7.

Python2 is no longer supported, and mixed environments in some systems are not desired (i.e. conda).

iannellog commented 4 years ago

I have planed to update the scripts for Python3. As soon as they are available I will add them to the distribution, but it will take a while because I am currently busy in other tasks.

iannellog commented 4 years ago

Chris, I have tested these scripts with Python 3.4.3 which is the version I have installed at the moment. Could you please check if they work with Python 3.7? Thanks a lot.

ParaTools_scripts.zip

chrisroat commented 4 years ago

I found one python2-ism that I hit in the "--fixed-tiling" branch, which I solved with the following change:

902c902
<       k  = size / wb
---
>       k  = size // wb