Hierarchical Popari approach

[alam-shahul]

This feature is critical to the utility of Popari in settings where the expression data is sparsely distributed in Euclidean space. It includes an implementation of:

• Spatial binning of sparse data
• Multi-resolution, hierarchical propagation of learned parameters
• An optimization scheme for superresolving embeddings using parameters learned at lower resolutions

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[alam-shahul]

Should probably add convenience functions for creating plots at different resolutions...

[alam-shahul]

Must also add a hyperparameter to control the balance between NMF reconstruction and bin reallocation during superresolution!