I use Python for everything. I don't have Stata anymore, I never know when something simple will be code I want to build on, and I code very quickly in Python.
R is much better for day to day data science than Python. Python is better if I need to program something like an optimization routine or a structural model