Predicting depressive and manic episodes in patients with bipolar disorder using statistical process control methods on passive sensing data
This repository contains the code for data processing, optimization and creating figures, written in Julia. It is intended to serve as a non-ambiguous documentation of our approach and as a starting point or reference for data processing in future studies in this research area.
Here is a quick overview of various files and folders of interest:
- Project.toml and Manifest.toml: These two files list the required Julia packages, their dependencies and the versions used in this project.
- Data.toml: This file, along with DataToolkit.jl, is used to organize and manage our data and contains, among other things, descriptions, checksums and references to the code that was used to create the datasets.
- src: This directory contains relatively general functions that could easily be reused in other projects. For example, to download and process data from movisens, calculate classification metrics or create plots for statistical process control.
- scripts: Contains specific code for loading, transforming and saving data, optimizing parameters and creating figures.