Hi again, Thank you for the suggestions!
I found the linked discussion very useful, and your notebook example using a mixture_density function was exatly what i was looking for.
I am hoping you can help me clearify one thing, @junpenglao:
I was testing for robustness of different magnitudes of the bin heights, so i gave a histogram with density=true as my data, essentially scaling down the bin heights. The model is not able to approximate this, nor is it able to find a fit when i scale the bin height by lets say 100. I have tried changing my w-prior to compensate for this, but I am getting very weird results.
Is this something that you can explain?
In advance, thank you so much for answering!