Hey! I’m building a Stick Breaking process using pymc guide. I have 2 dimensional data that correspond to metrics about vehicle utilization. The model fits correctly for the 2 variables under study:
I’m at odds however about how to translate this to 2D:
Here’s the first attempt at displaying the categories using the distance to the calculated mean of the component (I’m just ordering the results and matching components that way)
And the components:
Unfortunately, this doesn’t quite capture the calculated proportions nor the variances of the model.
Is there a better way to colour this graph?
Thanks!!