kutaslab / fitgrid

Multichannel event-related time-series regression modeling for EEG, MEG, and sensor array data
https://kutaslab.github.io/fitgrid
BSD 3-Clause "New" or "Revised" License
8 stars 4 forks source link

[JOSS] saving and loading grids: HDF5 #199

Open sappelhoff opened 3 years ago

sappelhoff commented 3 years ago

(related to openjournals/joss-reviews#3293)

I stumbled over this warning on loading and saving "grids" in the docs:

Fitted grids are saved and loaded with Python pickle which is not guaranteed to be portable across different versions of Python. Unpickling unknown files is not secure (for details see the Python docs). Only load grids you trust such as those you saved yourself. For reproducibility and portability fit the grid, collect the results you need, and export the dataframe to a standard data interchange format.

Wouldn't it be relatively straight forward to write a function that takes a "grid", saves it to hdf5, and loads it back from hdf5 on demand? Given that hdf5 is already a dependency of this project, this seems like a low hanging fruit to me that would greatly improve usability of the project.

turbach commented 3 years ago

@sappelhof I'm tagging this as an enhancement, but I think there is more to it. I'll leave the issue open for discussion if there's interest.

sappelhoff commented 3 years ago

okay, fair enough: out of curiosity: What are the reasons you think "there is more to it"?