umap icon indicating copy to clipboard operation
umap copied to clipboard

Looking for most efficient way to run transform()

Open Alex-EEE opened this issue 2 years ago • 0 comments

We're storing the results of umap.fit by pickling the umap object. I see this saves all the training ("raw") data, which is creating files close to 1 GB for our training data. Is there a way or mode to only save something much smaller, like just the embedding, or just the reduced training data, etc, to enable calling transform() with new samples?

Alex-EEE avatar Aug 31 '23 23:08 Alex-EEE