Dealing with unconsistent sfreq

[lkorczowski]

It seems that after refactoring all from 200 to 250Hz (#61 ), the problem was that half of the recording are 200Hz and the other half to 250Hz.

We can either:

OPTION 1 not dealing with different sample rate (e.g. changing duration)**

pro:

• more scalable

con:

• number of samples per epochs will be different which can modify a bit the results

OPTION 2 re-sampling all to 200 Hz**

pro:

• all epochs for all subjects are consistent

con:

• computationally intensive
• non-proportional resampling interpolation could actually modify the statistical properties of the signals.

Anyway, there is no-clear best practice here but I actually prefer not to resample. Less struggle for OPTION 1.