As shown in the picture I am not sure what to include in observed value. My dataframe has an customer index and on the right side for each customer has sequence of transactions. Would like to fit normal distribution. Thanks in advance.
Its usually easier to use flattened array(s) for model input. If you aren’t doing anything complicated like hierarchical modeling where you care about what customer fed in what transaction sequence, then you can create a 1d flattened array of all the transactions and feed this array as the observed input.
1 Like