original experiment and experiment_id must be dropped when making s
so as to have a the same OD_mean for all experiment_ids.

Changing assignment of cvar fixed by conditioning on value of dvar.

refactored importdf

Added preprocess_data to omfitderiv; ignore all NaN in omplot

Deleted old code & fixed bug in sampling background; fixed bug in showwells

Merges df correctly; fixes to index of sc; added predicted_fl to s;

Standard deviations are estimated assuming a log-normal distribution
consistent with the rest of the code.

More flexibility to run correctOD.