You clearly haven't used pythons major data science packages. Scikitlearn and statsmodels are known as being specifically easy to use with a uniform API tha

You clearly haven't used pythons major data science packages. Scikitlearn and statsmodels are known as being specifically easy to use with a uniform API that disparate R packages lack(though this is somewhat fixed with caret). Further, for bayesian inference, Pymc 3 is more powerful and concise than training a model in stan.