Scene Segmentation on ISIC using RF

[amyvanee]

Reference issue

Continued progress on issue #39 on scene segmentation.

Type of change

Added

• tutorial notebook (scene_segmentation_rf_ISIC_tutorial.ipynb)
• tutorial functions (scene_segmentation_rf_ISIC_tutorial_functions.py)
• updated tutorial list (tutorials.rst)

Fixed black formatting of documents

• benchmarks/sims/spiral_sim.py
• proglearn/tests/test_system.py
• proglearn/cifar_exp/experiment_varying_task_sample.py

What does this implement/fix?

Added another tutorial on scene segmentation using rf on ISIC dataset and fixed formatting of pre-existing files

Additional information

Neuro Data Design II, Spring 2022

[codecov[bot]]

Codecov Report

Merging #536 (418944c) into staging (d246c60) will not change coverage. The diff coverage is n/a.

@@           Coverage Diff            @@
##           staging     #536   +/-   ##
========================================
  Coverage    90.09%   90.09%           
========================================
  Files            7        7           
  Lines          404      404           
========================================
  Hits           364      364           
  Misses          40       40           

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update db0c6f8...418944c. Read the comment docs.

[amyvanee]

@amyvanee you need to also comment what other files you corrected to pass the style checks.

Updated my comment, thank you!

[jdey4]

I see 2 scene segmentation tutorials in docs/tutorial which should be in docs/experiment.

[amyvanee]

This file belongs to docs/tutorial, right?

Yes! They were from benchmarks/cifar_exp, proglearn/tests and proglearn/sims

[amyvanee]

I see 2 scene segmentation tutorials in docs/tutorial which should be in docs/experiment.

should these go under experiments instead? The notebooks we had previously were placed under tutorial?

[jdey4]

@amyvanee I would get rid of the "Dimensions of Images" blocks. The scatter plots are not informative. And why does the precision-recall curve show some dots instead of a continuous curve?

[jdey4]

I see 2 scene segmentation tutorials in docs/tutorial which should be in docs/experiment.

should these go under experiments instead? The notebooks we had previously were placed under tutorial?

@amyvanee they should be in the experiments folder. The location for the current PR is correct. But it was not correct for the PRs last semester.

[amyvanee]

I see 2 scene segmentation tutorials in docs/tutorial which should be in docs/experiment.

should these go under experiments instead? The notebooks we had previously were placed under tutorial?

@amyvanee they should be in the experiments folder. The location for the current PR is correct. But it was not correct for the PRs last semester.

Sounds good, thank you! I also fixed the precision-recall curve and removed the unnecessary information about the dimensions of images.

[amyvanee]

@amyvanee Check your website preview and fix the math equation. Other than that LGTM.

Fixed, thank you!!