I’m trying to make an out-of-sample prediction using a model_factory setup. But I’m getting a shape error when using pymc5, but the same code works on pymc3. The error is due to the size deference of the “seasonality” random variable, which assumes it needs to have the same size as the training data…

My hunch is that because you are creating a new model, and not using MutableData to store your fourier_features, PyMC does not know the shape has changed and does not resample the fourier_coefs. You can either make those MutableData, or you can manually add fourier_coefs (and your observed variable)…

Thanks @ricardoV94 :slight_smile: You are certainly the shapes master

Out of sample prediction

Questions v5

RavinKumar April 7, 2023, 3:38am 22

Thanks @ricardoV94 You are certainly the shapes master

Topic		Replies	Views
Shape error when making out-of-sample predictions version agnostic shape_issue , prediction	13	516	January 10, 2024
Pymc5 out of sample v5 modeling	9	1364	August 20, 2023
Issues when trying to do out of sample prediction v5	2	180	April 9, 2024
Sample_posterior_predicitve not catching shape of new data v5 prediction	10	1372	August 24, 2022
How to predict new values on hold-out data Questions	24	13633	July 22, 2020

Out of sample prediction

Related topics