Yes, true makes sense, thanks. Yet the problem remains that with
Initial evaluation results:
{'sigma': -1.07, 'pred_err_scale': -1.66, 'log_level_scale': -2.02, 'combined': -263.89, 'respone': -inf}
could it have something to do with the line p = pm.logcdf(combined, x)? That internally there is a problem with estimation?