Stan is a platform for probabilistic programming. To demonstrate its features I did data analysis of wind energy capacity factor in Finland. Wind energy is feasible in Finland, and we have quite high seasonal variance, so modeling wind data makes an interesting case. This case study presents a Bayesian data analysis process starting from data, modeling, model diagnostics to conclusions.

Statistical modeling on a modern computing platform such as Stan let’s you construct the model quite freely. I mean, you can all but ignore such constraints as conjugate priors. Stan’s implementation of Hamiltonian Monte Carlo can generate reliable estimates of very hard integrals. You can pay more attention to the model at hand, instead of computational constraints.

The full report is here: Wind Power Generation Efficiency and Seasonality. One reviewer of the report said it well: “In many cases, modeling itself only produces more development ideas than the answers themselves, which is also very evident in this work.”

Original data is from Fingrid (data.fingrid.fi, license CC 4.0 BY).