Hi Jacob,

Thank you for you feedback. I almost agree with everything you said (almost because I still don't see how I could fit my problem of understanding price elasticity of demand into a logistic regression given that my target is the price).

The thing is that of course, such a task can be extremely complicated (up to the degree you want to go to). You can go bayesian indeed. You can also use some more econometric oriented solutions like mixed effects. You can also use some more traditional statistical models like generalized linear models...

There can be a huge amount of modeling work, but... for the context of this very mission... it was not needed. My boss at the time gave me a bone because I expressed my frustration (it was an internship, I wanted to learn as much as I could). I worked 5 full months on this task and performed great analysis. But in the end it was not needed. The client did not give a dime about it. In the context of the client mission, we should have stopped at performing a simple linear regression on the aggregated monthly data to be able to express the linear regression coefficient as the overall general price elasticity of demand. The client did not care any further.