I ran a gaussian mixture model with 7 clusters on my data. My data has been PCA transformed with 200 components. Then I extracted the means of each cluster and applied the predict_proba function on those cluster means. And this is what I get. Each row is a cluster center, and each column is the probability of this point belong to each cluster
I thought the diagonal of this data frame should be 1, as each cluster mean should have the highest probability of belonging to its own cluster. I'm wondering why some cluster means have been assigned to a different cluster?