Different results between Jupyter Notebook and Readme

[Steven-Livingstone]

First of all AWESOME STUFF!

However, in an effort to reproduce your results locally I spotted a few differences between your Jupyter notebook results and the ones given in the readme.

Value	stan_mmm.ipynb	readme.md
MAPE	26.06330645337451	8.63
RMSE (log-log model)	0.13647800963534104	0.04977
MAPE (multiplicative model)	20.313393234793644	15.71

Have the Readme results been generated using different priors/parameters, or am I missing something obvious?

[sibylhe]

Don't bother, it doesn't really matter. They were generated from different datasets. The dataset published is tweaked, and the Jupyter notebook is done on top of it.

[Steven-Livingstone]

Ah, that makes sense then.

Thanks for the quick clarification. :)