STAR Digital Phenotyping Pipeline

The file tasks.py organizes the tasks that can be completed by these scripts using the invoke library. The file invoke.yaml specifies some variable values used by tasks.py.

In order to run invoke, you need to first load the python module, and then load the star_diph conda environment (which must be created for your user using the star_diph.yml file.

Example environment creation

module load python/3.10.9-fasrc01 
mamba env create -f star_diph.yml

Example environment activation and invoke

The command, invoke -l below will list the tasks. To get help for a task, e.g., the build task, you can run invoke -h build.

module load python/3.10.9-fasrc01
mamba activate star_diph
invoke -l
invoke -h gps

Other necessary info

In order to work properly, these scripts need an updated csv taken from the STAR Participant Tracker Google Sheet which should be saved in the same directory as this README file and named STAR_Month-Day_Subject-Calculator.xlsx - Sheet1.csv.

In order to make sure that the script is using the most up-to-date phone and call logs, please run

invoke sync-phone-data

To build all data and reports, run

invoke build -a

This will submit multiple sbatch jobs.