I know that to reproduce results in pymc, you have to set the random_seed when sampling.
However, in pymc-marketing DelayedSaturedMMM, when I fix the random seed when fitting the model and when predicting, I always get new results.
I tried to set the random_seed here:
mmm.fit(X=X, y=y, target_accept=0.85, chains=4, nuts_sampler="numpyro", random_seed=42)
and here:
mmm.predict(X=X, random_seed=42)
And I still get different results every time.
New contributor
Radhouane Hamroun is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.