Find images #54
Find images #54
Conversation
|
@jennyfolkesson I see that the error in well workflow came because of refactoring of the image io methods. It is great that you started to refactor and are now adding tests in pytest format. Please implement the following to make sure changes from this branch are useful in master:
|
Codecov Report
@@ Coverage Diff @@
## master #54 +/- ##
=======================================
Coverage 22.36% 22.36%
=======================================
Files 21 21
Lines 1359 1359
=======================================
Hits 304 304
Misses 1055 1055
Continue to review full report at Codecov.
|
|
@mattersoflight I've rebased wrt master and checked that it works for array_fit and array_interp. Please take a look again. Thanks! |
array_analyzer/utils/io_utils.py
Outdated
| for ext in extensions: | ||
| search_str = os.path.join(input_dir, '**', ext) | ||
| image_names.extend(glob.glob(search_str, recursive=True)) | ||
|
|
There was a problem hiding this comment.
searching the folder tree recursively returns too many images, and assigns wrong images to the wells.
There was a problem hiding this comment.
Above screenshot was taken with following arguments:
-e -i="/Volumes/GoogleDrive/My Drive/ELISAarrayReader/images_scienion/2020-04-08-14-48-49-COVID_concentrationtest_April8_images" -o="~/Documents/images_local/2020-04-08-14-48-49-COVID_concentrationtest_April8_images" -wf=array_fit
There was a problem hiding this comment.
When I tested it on the data paths I didn't get these issues, but I never write outputs in the same directory as the input so that's probably why I missed that. Thanks for not only catching it but also fixing it!
There was a problem hiding this comment.
When I tested it on the data paths I didn't get these issues, but I never write outputs in the same directory as the input so that's probably why I missed that. Thanks for not only catching it but also fixing it!
Somewhat related -- do we want to generally allow writing reports and outputs to the same directory as the raw data files? It does cause problems when, for example, searching for excel metadata spreadsheets.
|
I've fixed the bug by restricting the filename search to either the input folder or one level down from input folder. The well workflow runs fine now and array_fit workflow runs too, but I don't see an output OD file. @jennyfolkesson please take a look. |
|
@mattersoflight I did get an OD file, but I named it similar to the debug excel files. I've now renamed the OD excel file to match the one in the interpolation workflow: 'python_median_ODs.xlsx'. I hope that helps. |
|
I changed the other two filenames to be consistent too. Merging. |
Added support to read either a directory containing images or a number of subdirectories containing one image each.
I also added tests to the code, which I recommend doing moving forward for those who are so inclined. It really helps if the repo is intended to have many collaborators.