I have a dataframe with 1M rows in which one column represents the log posterior probability. Many of t