numpy is excellent, actually

docs can be a little hard to find exactly what you want, but the info is all there. no complaints about the library itself once you know what you are doing

matplotlib hasn't totally clicked in the same way. the switching between the functional and OO interfaces doesn't really help

@koopa I think you can avoid the functional part if you want
One issue matplotlib has to new people is it emulates MATLAB, so if you don't know (lucky you) MATLAB its syntax is less coherent

It'd be neat to get a more pythonic wrapper over matplotlib (because the plots it makes do look great), maybe one exists

@koopa numpy is amazing, so many times have I wondered if I can do X in numpy and it works exactly as I expect