if I have bimodal data, and I want to transfer it to two mixture distributions, which transformation method could I use?
I have one-dimensional data. I plot the histogram for it. It seems that it has two modes. I applied the Gaussian mixture model (GMM) but the q-q plot is not very good. I am planning to transfer data by using logit transformation because my data is probability and greater than 0 but less than 1. however, after transformation, the histogram became unimodal. I think that there are specific transformations for the mixture (bimodal) data