Authors: Ishita Jain, Mihir Gadgil

Read the report.

The data can be obtained from Kaggle. Extract it into the data directory.

Running these script requires a file with some utility functions. It can be obtained from the webpage of the Doing Bayesian Data Analysis book by John K. Kruschke.

The generate_mcmc.R script can be used to create the MCMC chains.

You will need to modify the script with the names of the models.

For the high rating model:

1. Change line 11 to yName = "Rating_high"
2. Change line 12 to fileNameRoot = "appRatingFilesHigh"

For the medium rating model:

1. Change line 11 to yName = "Rating_med"
2. Change line 12 to fileNameRoot = "appRatingFilesMed"

For the low rating model:

1. Change line 11 to yName = "Rating_low"
2. Change line 12 to fileNameRoot = "appRatingFilesLow"

We could write this in the script itself, but that means the MCMC simulations would run in sequence one after the other. This takes too much time.

Whereas you can run them in parallel if you want to by manually editing the script.

The chains can be analyzed and the posterior distributions can be plotted using the plotting.R script.

We have included the results of our MCMC simulations in the repository. You can directly analyze the results without the need of generating new MCMC chains.