Possible issue with the way trials are balanced in classifs and ttests

[hyruuk]

For now, different trials are retained for each test (each electrode). It might be important to use the same trials for all electrodes.