We're using the World Happiness Report from 2018. We made some preprocessing to the data so that it is easier to work with. If you want to see the processing steps, check this file. The resulting data frame is called preprocessed_data.csv.
A short exploratory data analysis with a description of important variables can be found here.
In case this is your first time trying out some Bayesian analysis frameworks, please make sure to have any frameworks you would like to try installed beforehand! The most popular languages are Stan and PyMC3 where Stan has a wide variety of interfaces to common programming languages
PyMC3: https://docs.pymc.io/
Stan has interfaces for R, Python, Matlab, Julia, Stata, Mathematica, and Scala. The most commonly used interfaces for Stan are R and Python. It also has 2 high-level interfaces for R that are very beginner-friendly: rstanarm and brms.
For use in R, consider also the complentary packages shinystan, bayesplot and loo.
https://mc-stan.org/users/interfaces/
If you use PyStan or PyMC3, consider also the package ArviZ. It is a Python package for visualization of Bayesian models https://arviz-devs.github.io/arviz/
Provided you have a local Anaconda installation (see here and here), you can use our prepared PyMC3/ArviZ environment (YAML-File).
This should come with the most relevant PyMC3-centric Bayesian Data Analysis packages. Simply run the following command
conda env create -f berlinbayesians_environment.yml
A short working example of how to implement a linear model in PyMC and visualize it with Arviz can be found here.
If you don't know where to start, consider some of these examples:
PyMC3: https://docs.pymc.io/nb_examples/index.html
Stan: https://mc-stan.org/docs/2_18/stan-users-guide/linear-regression.html
PyStan: https://mc-stan.org/users/documentation/case-studies/pystan_workflow.html
RStan: https://mc-stan.org/users/documentation/case-studies/rstan_workflow.html
RStanArm: http://mc-stan.org/rstanarm/articles/continuous.html