scvi-tools
scvi-tools copied to clipboard
scverse
Deep probabilistic analysis of single-cell and spatial omics data
There were sources of confusion among more than one users about how the model consumes this, so this is adding some more context to the docstrings to provide clarity on...
The issue here is that PyTorch Lightning uses its wrapper of an optimizer to track steps. It uses post step hooks to increment the global step of the trainer. Here...
the `get_imputed_values` method from gimvi not working? ```python from scvi.external import GIMVI GIMVI.setup_anndata(adata_sp_train, layer="counts") GIMVI.setup_anndata(adata_sc, layer="counts") model = GIMVI(adata_sc, adata_sp_train) model.train(10, batch_size=256) imputed = model.get_imputed_values(normalized=True, batch_size=256) ``` ```pytb --------------------------------------------------------------------------- ValueError...
updates: - [github.com/PyCQA/flake8: 5.0.2 → 5.0.4](https://github.com/PyCQA/flake8/compare/5.0.2...5.0.4)
![pre-commit-ci[bot] avatar](https://avatars.githubusercontent.com/in/68672?v=4 "pre-commit-ci[bot] avatar"){kind=avatar}
Bug: When using `gene_likelihood!="zinb"`, the function `get_likelihood_parameters()` would fail. To reproduce: ``` model = SCVI(adata, gene_likelihood="nb") model.get_likelihood_parameters() ``` Fix: Added `if self.module.gene_likelihood == "zinb":` to a few lines.
Hi, I have met such a strange bug. The kernel always restarted when I ran sc.pp.neighbor after importing scvi package. ```python # code for illustration # without importing scvi package,...
Hi @adamgayoso and @PierreBoyeau. As we discussed recently via email, in our lab we were able to use the logic in [michael/autotune](https://github.com/scverse/scvi-tools/tree/michael/autotune) to do hyperparameter optimization. I've implemented the changes...
Mentioned by @jjhong922 in #1608 - need a Jax equivalent of the PyTorch FCLayers to reduce redundancy in FlaxEncoder/Decoder