# From Idiographic to Nomothetic Prediction and Back: Leveraging big data for real-time stress prediction Project hub for idiographic prediction project. Contains scripts used for analyses. Folder structure as follows: - **scripts:** Contains scripts used for analysis on the Leiden HPC cluster. Each python script was run using a .sh wrapper for the SLURM cluster. - `functions.py`: functions used for data processing, model building etc... - `dataprep.py`: script to do within-subject imputation on the data prior to analysis. Saves the final output so it can be used by the rest of the scripts. - `model_tuning.py + bash_tuning.sh`: scripts used to run the tuning process on 20 subjects to determine the ideal network architecture. - `model_par.py + bash_par.sh`: scripts to run the idiographic LSTMs using either offline or online architecture, on either the smartphone data (ema), passive sensing data (epa), or both (all). - `model_ensenmble + bash_run.sh:` script to run ensemble models with an 80/20 split.