Hierarchical Bayesian analysis is applied to model the response time for questions on Stackexchange websites.

• p7zip used in extract.sh to extract .7z files in Stackexchange dump. Any other package can be used as long as the 7za command in extract.sh is appropriately modified.
• Stan for MCMC inference.
• R interface for Stan: rstan
• Python interface for Stan: PyStan

Run extract.sh in the project directory to extract stackexchange dump. Individual .7z files are assumed to be located in stackexchange folder in the project directory, and the extracted files are stored in extracted folder. The unpacked directory, including the directory structure are automatically created by the script. Additionally, it copies the directory structure of unpacked folder to data folder, for use by parser.py. The directory names can be changed in extract.sh

Run parser.py to generate csv files containing relevant attributes. The generated csv files are stored in data directory. The directory name can be changed in extract.sh and parser.py

• Ordinary Least Squares Regression: Implemented in linear-regression.R
• Bayesian Linear Regression: Implemented in hierarchical-linear-regression.R, with the model specification in bayesian-regression-pooled.stan
• Hierarchical Linear Regression: Implemented in hierarchical-linear-regression.R, with the model specification in bayesian-regression-hierarchical.stan
• Generalized Linear Model: Using Exponential and Gamma link functions. Implemented in stan_model.py
• Probit Regression: Discretizing the regression problem. Implemented in stan_model.py