Unequal trial probabilities, VIF, correlations

[tilsley-p]

Hello,

I have been trying to see whether I could use this package to optimise an fMRI design (thanks first for taking over the upkeep and debugging for python3!), however I'm not so sure on the concept/results I get out.

1. When I put equal probabilities (6 conditions so 1/6 everywhere) I don't get an equal number of trials/condition output, so one condition may have e.g., 35 trials, and another 30 trials. Is there a way to force that to be equal?
2. When asking for uniform jitters, the output jitters also do not seem to be completely uniform.
3. Does the toolbox use or account for VIFs? When I create a randomised design myself the VIFs seem lower than with the design created by neurodesign (all are still less than 2). & When looking at the correlation matrix, the neurodesign design seems always to have values ranging from -0.1 to -0.2, whereas my random design has a range of -0.2 to +0.2, so was just wondering whether negative correlations are preferable to positive? or which from VIF/correlation/efficiency to prioritise?

Best,
Penny

p.s. The classes.py has an error I think in line 787 warnings.warns( should be warnings.warn(