Save a PaCMAP model

[chipsin66]

Hi, I am really enjoying your PaCMAP package. Getting some great results. I have a request for a new feature (unless this is a feature that exists and I am not aware of it).

I am running PaCMAP on large datasets and then using the model with the transform function to transform new data into the existing embedded space. With these models developed using large datasets it would be great if I could save a PaCMAP model locally and then reloaded it and apply it to new data, rather than having to run the PaCMAP model again when I start a new session. An analogy would be I often save my XGBoost models and apply them to new data rather than rerun the model each time I have new data.

Is this feature already available or something that can be implemented?

Thanks again.

[hyhuang00]

Thank you for your interest in PaCMAP and your suggestion! We do have implemented methods to save PaCMAP locally, but as reported in #34 there seems to be some problem when saving the ANNOY trees. We will get it fixed soon and let you know!

[hyhuang00]

V0.7.0 has been released on GitHub, which comes with the save and load feature. I will upload the version to PyPI later today. Please let us know if the new version solves your problem!