radius-solaris

This project is a research endeavor into developing a linear regression model for predicting solar radiation. What we find is that a linear regression model is not particularly well equipped for such a task because it requires a more complex model. We eventually reach that conclusion because although we find that our model can explain 65.3% of variability observed in solar radiation, it does not uphold one of the most important assumptions in linear regression, which is the constant variance assumption. Thus, we advise against using linear regression to predict solar radiation due to the temporal nature of solar radiation.