Reusing calculated data for gene trend models

[jmkotah]

...

Hello again,

I wanted to ask a question to do with the pl.gene_trends function.

I was wondering how the fitted model can be reused for subsequent line plots. From the docs I see that there is an option to set return_models=True, but I am not sure how to load it back for the plotting function. I have tried saving the resulting model dictionary as a variable, and using that dictionary in the model argument, i.e.:

a = cr.pl.gene_trends(ad, model = model, ...) #where model is the GAM model as in the vignette
cr.pl.gene_trends(ad, model = a, ...)

But this results in recalculating the gene trends before plotting anyways. Do you have advice on how to implement this properly?

Thanks!