Some distributions only fit positive data

[helen-zhu]

In fit.and.evaluate, we include options for weibull, gamma, exp and lnorm which are distributions that only have data density for positive values. This is counter-intuitive given that we generate our distribution of aggregated z-scores by summing up negative z-scores.

[helen-zhu]

I think it should be noted that in the functions cosine.similarity.cutoff and cosine.similarity.iterative, the signs of the data are flipped before fitting distributions. We should probably add a parameter to fit.and.evaluate to match to avoid confusion.

[helen-zhu]

I think it might actually just be the zeros that are throwing this error.