This repository contains the codebase for the PREACT-digital project, associated with the paper:
Towards JITAI – Moment-to-moment Prediction of Negative Affect in Internalizing Disorders using Digital Phenotyping
-
model_pipeline/
Contains all model definitions, training scripts, and evaluation code related to the predictive modeling of negative affect. -
src/
Houses preprocessing functions and utilities required to transform raw data into a format suitable for model training and evaluation. -
relevant notebooks/
- X1_Short_Term_Prediction.ipynb
- Notebook for running the model pipeline and create plots. It utilizes modules in
model_pipeline/and processes the output from the preprocessing step.
- Notebook for running the model pipeline and create plots. It utilizes modules in
- 03_JITAI_Preprocess.ipynb
- Notebook for executing data preprocessing steps, aligning EMA with passive data and aggregating them. It relies on functions from
src/to clean and prepare raw data.
- Notebook for executing data preprocessing steps, aligning EMA with passive data and aggregating them. It relies on functions from
- 01_Data_Preprocess.ipynb
- Notebook for basic data preprocessing, including concatenation of EMA and passive data with study monitoring data, and timestamp alignement.
- X1_Short_Term_Prediction.ipynb
