The scripts in this repository requires data from the Tromsø surveys 4-7 in long format (vertically stacked).
To run the scripts that generates the analysis, navigate to setup and run runMeAtStartup.m, which adds folders essential folders to path, and loads the data and processes the data and stores it in a workspace variable called data. The expected folder structure is as follows
├───data
│ ├───diary data
│ └───fitbit data
├───saved matlab files
│ ├───app config
│ ├───models
│ ├───saved variables
│ ├───tables
│ └───user-tables
├───src
│ ├───¤¤ article
│ └───setup
The data folder is expected to contain the data: hscl_long.csv and
hscl_wide.csv (the latter is not nessecary for the main analysis). The matlab app can be found in src/prototype/secondPrototype.mlapp. Before running the app, the script src/¤¤ article/model_regression.mat must be run, which fits and dumps the linear model into the saved matlab files/models folder.
To produce a result in the article, navigate to the src/¤¤ article folder, and run the corresponding script contained in that folder.