R code for Panel Data Analysis with Expectation Maximization iteration for missing values. Final models are heteroskedastic, so Robust Covariance Matrix is used. Since the panel analysis didn't produce any potential fixed effect models, Chow Test wasn't included.
Data used is total energy use and average temperature in each province in Indonesia, both from 2011 to 2015.
note: this is added for archival purposes. Original version is runnable, but I give no guarantees that code is runnable considering code may be accidentally or purposely edited without reverting to original state after first use