-
Notifications
You must be signed in to change notification settings - Fork 2
Helper Page | Jamboree DeepProfiler day
Michael Bornholdt edited this page Nov 18, 2021
·
9 revisions
DeepProfiler repository Cytominer-eval Pycytominer
- DeepProfiler Wiki is WIP. Please comment your feedback directly in the document!
- Error log
Download this example data! (Ignore Virus error) Unzip file and move to to a convenient location. This is your project folder!
Example self-trained model: Move this to /project/outputs/profile_folder/checkpoints/
Pull with
docker pull michaelbornholdt/deep_profiler:jamboree
Run with
docker run -it --rm -v ~/project_folder:/project_folder michaelbornholdt/deep_profiler:jamboree
Run Aggregation with code example found in the Wiki
Reading metadata form /git/tests/example_data//inputs/metadata/index.csv
2021-11-18 15:35:57,197 - INFO - Total plates: 2
/git/tests/example_data//outputs/intensities//Week1_22141/Week1_22141.pkl exists
/git/tests/example_data//outputs/intensities//Week1_22123/Week1_22123.pkl exists
Illumination complete!
Reading metadata form /git/tests/example_data//inputs/metadata/index.csv
2021-11-18 15:35:57,320 - INFO - Total plates: 2
Percentiles for the lower_percentile >> DNA : 304.0 Tubulin : 736.0 Actin : 368.0
Percentiles for the upper_percentile >> DNA : 7728.0 Tubulin : 41888.0 Actin : 20544.0
Percentiles for the lower_percentile >> DNA : 288.0 Tubulin : 752.0 Actin : 352.0
Percentiles for the upper_percentile >> DNA : 7696.0 Tubulin : 47520.0 Actin : 18768.0
Progress |##################################################| 100.0% Complete
Progress |##################################################| 100.0% Complete
Compression complete!
Running profiling will throw a bunch of warnings - please ignore :)
Extracting output from layer: pool5
Week1_22141/B02-s2 (240 cells) : 2.00 secs
Week1_22141/B03-s2 (56 cells) : 0.27 secs
Week1_22141/B02-s1 (276 cells) : 3.82 secs
Week1_22141/B04-s1 (90 cells) : 0.45 secs
Week1_22141/B03-s1 (103 cells) : 0.42 secs
Week1_22141/B04-s2 (104 cells) : 0.44 secs
Week1_22123/B03-s1 (112 cells) : 0.48 secs
Week1_22123/B04-s1 (89 cells) : 0.36 secs
Week1_22123/B03-s2 (28 cells) : 0.13 secs
Week1_22123/B04-s2 (79 cells) : 0.34 secs
Week1_22123/B02-s2 (149 cells) : 0.69 secs
Week1_22123/B02-s1 (206 cells) : 1.04 secs
Profiling: done
Aggregating features in /git/tests/example_data/outputs/pre/features
Week1_22141/B02
Week1_22141/B03
Week1_22141/B04
Week1_22123/B03
Week1_22123/B04
Week1_22123/B02
!
- download example data to
/jamboree/
unzip example_data.zip
docker run -it --rm -v ~/jamboree:/jamboree michaelbornholdt/deep_profiler:jamboree
- check that the mount worked:
ls /jamboree/example_data/
should give inputs and outputs python3 deepprofiler --root=/jamboree/example_data/ --config=config.json prepare
- in
/jamboree/example_data/inputs/config/config.json
change the profile: checkpoint to"None"
python3 deepprofiler --root=/jamboree/example_data/ --config=config.json profile
- create aggregate.py in /example_data/ with code above and run it.
- DONE! :)