Assignment 3

Prerequisites for running queries:

  1. Python (3.7+)
  2. Pytest
  3. Pytorch
  4. Nibabel

Input Data

Queries of Assignment 3 expect input files and directories as follows :

  1. data directory : "data" consists of "datasets" and "output" directories

(output directory is created when tasks are executed)

  1. datasets directory : consists of "seg" directory, the npy files for all the MRIs, cnn_best.pth file and the ADNI3.csv metadata file.

(This is the directory where the bg_file and test_file are created during task 1)

  1. output directory : will consist of SHAP directory and task-1.csv file.

  2. SHAP directory : will consist of data and heatmaps directories.

  3. SHAP/data directory : consists of the shap_value npy arrays from task 2.

  4. SHAP/heatmaps directory : consists of the 2 heatmaps generated from task 3.

  5. skeleton directory : consists of the data_util.py file, the explain_pipeline.py file, the model.py file and the Assignment3_test.py file.

Running queries of Assignment 3

Example command :

python explain_pipeline.py --task 3 --dataFolder ../data/datasets/ADNI3 --outputFolder ../data