Compute the correlations between the outputs of the TrainingCenter.
Feel free to start a discussion to ask anything here
To setup the virtual environment:
pip install -e .It fetches the data from TrainingCenter, adds to it the predictions and the feature importances from the dumps and compute the residuals. The outcoming data is stored at data/main_category/category.feather.
To launch this step:
./residual/shell_script/run_residual.shIt uses the residuals from age prediction computed in the last step as a predictor for survival along with the age, the sex and the ethnicities. It outputs the results on a google sheet.
To launch this step:
./log_hazard_ratio/shell_script/run_log_hazard_ratio.shIt stores the feature importances, that were gathered in the first step, at data/feature_importances/main_category/category.feather.
To launch this step:
./feature_importances/run_feature_importances.shIt computes the correlations between the residuals and between the feature importances from the data that has been stored in the first step. The outcoming correlations are stored at data/correlation/residual/ and data/correlation/feature_importances/.
To launch this step:
./correlation/run_residual_correlation.sh
./correlation/run_feature_importances_correlation.sh┣ 📦TrainingCenter
┃ ┗ 📂data
┃ ┣ 📂examination
┃ ┃ ┗ 📜[category].feather
┃ ┣ 📂laboratory
┃ ┃ ┗ 📜[category].feather
┃ ┗ 📂questionnaire
┃ ┗ 📜[category].feather
┣ 📦CorrelationCenter
┗ 📂[...]
📂data
┣ 📂correlation
┃ ┣ 📂feature_importances
┃ ┃ ┣ 📜pearson_[...].feather
┃ ┃ ┣ 📜spearman_[...].feather
┃ ┗ 📂residual
┃ ┣ 📜number_participants_[...].feather
┃ ┣ 📜pearson_[...].feather
┃ ┗ 📜spearman_[...].feather
┣ 📂examination
┃ ┗ 📜[category].feather
┣ 📂feature_importances
┃ ┣ 📂examination
┃ ┃ ┗ 📜[category].feather
┃ ┣ 📂laboratory
┃ ┃ ┗ 📜[category].feather
┃ ┗ 📂questionnaire
┃ ┃ ┗ 📜[category].feather
┣ 📂laboratory
┃ ┗ 📜[category].feather
┗ 📂questionnaire
┗ 📜[category].feather