Danielsf/dev/downsampled/video

[danielsf]

This PR adds a utility to generate downsampled video files (.avi, .mp4, etc.) for those who do not want to use the .webm files currently output by the motion correction pipeline. There is a module (single_video) for generating a single downsampled video and a module (side_by_side_video) for showing two videos side by side (e.g. pre- and post-motion correction). Examples of the side by side video module's output can be found here

/allen/aibs/informatics/danielsf/downsampling_example

This work is not associated with any specific ticket (unless maybe #412), but having this tool will be necessary to enable the science team to evaluate our prospective motion correction configurations.

[codecov]

Codecov Report

Merging #415 (3602fde) into main (272c74a) will decrease coverage by 0.14%.
The diff coverage is 89.21%.

❗ Current head 3602fde differs from pull request most recent head 8c76d54. Consider uploading reports for the commit 8c76d54 to get more accurate results

@@            Coverage Diff             @@
##             main     #415      +/-   ##
==========================================
- Coverage   92.47%   92.32%   -0.15%     
==========================================
  Files          84       88       +4     
  Lines        5182     5423     +241     
==========================================
+ Hits         4792     5007     +215     
- Misses        390      416      +26     

Flag	Coverage Δ
event_detection_tests	30.42% <24.89%> (-0.26%)	⬇️
general_tests	90.09% <89.21%> (-0.05%)	⬇️
not_container_tests	90.15% <89.21%> (-0.05%)	⬇️
suite2p_tests	34.27% <24.89%> (-0.44%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/ophys_etl/modules/downsample_video/schemas.py	72.34% <72.34%> (ø)
...ophys_etl/modules/downsample_video/single_video.py	81.25% <81.25%> (ø)
...etl/modules/downsample_video/side_by_side_video.py	82.35% <82.35%> (ø)
src/ophys_etl/modules/downsample_video/utils.py	95.65% <95.65%> (ø)

[aamster]

Please add the same description here as in the code, which is better

        Factor by which to speed up the movie *after downsampling* when writing
        to video (in case you want a smaller file that can be played back
        faster)

[aamster]

I think this package should just be called video, as it's possible to apply no downsampling but still create a video

[aamster]

SideBySideVideoSchema

[danielsf]

The python 3.7 build keeps timing out on CircleCI on tests that this PR does not touch. CircleCI reports itself as in a degraded state. I'm going to merge now on the assumption that that is our problem. If the problem does not resolve itself tomorrow, we can look into it.