On Thu, Sep 25, 2014 at 1:06 PM, Matt Davis <[email protected]> wrote: > Pandas is becoming so mature and popular now that I think there are whole > swaths of new Python users that have no exposure to raw NumPy. But at the > same time there are plenty of people who just don't need Pandas and for whom > teaching learning Pandas is not going to be helpful when they go back to > work. Might this be a situation where it's worth maintaining parallel sets > of lesson material with and without Pandas?
I think this could definitely be the case here. I agree with April and Matt's points regarding why Pandas is valuable. One additional point is that standard Numpy arrays don't handle mixed data types. So if we have data that includes both numbers and strings (extremely common in my world) then you have to move to a Numpy structured array. These have even great complexities with getting imports to work and when you do they work like a somewhat awkward version of a DataFrame. That said, we've definitely had feedback from multiple folks that Numpy is better for the folks that they are instructing, and as Matt points out when you actually want to use a proper 2D array, then Pandas is useless. There are lots of areas of science that are very data table focused, but also lots of areas that are very 2D array/matrix focused, so I think there's a good argument for parallel lessons. Ethan _______________________________________________ Discuss mailing list [email protected] http://lists.software-carpentry.org/mailman/listinfo/discuss_lists.software-carpentry.org
