How do you use the augmented matrix from harmony_timeseries in diffmap?

Hi there,

I was wondering if anybody knows an easy/seamless way to use the augmented MNN matrix from running harmony_timeseries in creating a diffusion map using diffmap?

I ran:
sc.external.tl.harmony_timeseries(adata_u, tp=‘MUT_WT’, n_components=None)

Which created:
harmony_aff_aug

I see that I would need to run sc.tl.diffmap using a different value for the “neighbors_key” parameter. Would I need to hard-code a changed value for: .uns[‘neighbors’]
.obsp[.uns[neighbors_key][‘connectivities_key’]]
.obsp[.uns[neighbors_key][‘distances_key’]]

If so, how would I go about hard-coding these changes if so?