Writing unit tests for prediction intervals

[arvganesh]

@ababii @vollmersj

I'm trying to determine a good way to test the code I wrote to compute prediction intervals for the models I've implemented. Do you have any suggestions?

[ababii]

@arvganesh @vollmersj You can compare it to the output of similar R or Python functions using one of the datasets that we discussed. Note that intervals have to be adjusted by the h-step standard deviation if you assume the N(0,1) distribution for errors. Could you also commit the ETS code, please?