Script execution sequence

The sequence of our codes and its arguments can be found in the commands_list.sh file. The following list indicates the different phases and its scripts.

0. Getting data

Script for internal use (Machine Learning in Mental Health Lab only).

1. download_datasets.py
2. combine_sites_data.py

1. Preprocessing

1. clean_biobank1_data.py
2. clean_clinical_data.py
3. demographic_balancing_adni_data.py
4. demographic_balancing_tomc_data.py
5. demographic_balancing_oasis1_data.py

2a. Bootstrap analysis

Analysis where we measure the performance of the normative approach using a bootstrap method. These scripts obtain the results from sections 3.1, 3.2, 3.3 of the paper.

1. bootstrap_create_ids.py
2. bootstrap_train_aae_supervised.py
3. bootstrap_test_aae_supervised.py
4. bootstrap_group_analysis_1x1.py
5. bootstrap_create_figures.py
6. bootstrap_hypothesis_test.py

2b. Classifier analysis

Analysis where we measure the performance of the binary classification (using RVM). To estimate the performance of the model, we used the .632+ bootstrap method. These scripts obtain the results from sections 3.4 of the paper.

1. classifier_create_ids.py
2. classifier_train.py
3. classifier_group_analysis_1x1.py
4. bootstrap_normative_vs_classifier.py

2c. Miscellaneous

1. univariate_analysis.py