search

mclaughlin6464 / pearce

A (better structured) package that implements the calculation of mock cosmological observables and their emulation.
MIT License
0 stars 1 forks source link

Consider turning training data into Pandas frames #13

Closed mclaughlin6464 closed 6 years ago

mclaughlin6464 commented 6 years ago

This will make slicing/cuts make more sense, and possibly other functions. However, may be slower and make playing with George harder. I will need to reboot how the data is structured when I move to cosmology, so this should also be considered during that development.

mclaughlin6464 commented 6 years ago

Training data are now hdf5 files. Though not pandas frames, which still have some external advantages, I think what I've done is satisfactory