v2.0.4

Fixed dcor import bug as an external dependency

v2.0.3

Bugfix for pp.pseudobulk when obs columns were not categorical.

v2.0.2

Small patch with the following bug fixes:

• Added ipywidgets as dependency
• Silenced xgboost warnings
• Handled corner case in bm.metric.auc when scores are all 0
• Fixed error in bm.metric.hmean when metrics were str instead of list
• Fixed error when obs col is list in dc.pp.pseudobulk
• Allowed ordering functions (dc.pp.bin_order, dc.pl.order and dc.pl.order_targets) not to be bound between 0 and 1

v2.0.1

Small patch to fix an error in decoupler.pp.pseudobulk when handling empty samples or features

v2.0.0

Major refactor to follow the scverse template.

decoupler 1.6.0

What's Changed

• Benchmark dev by @demian1 in #8
• Merge benchmark devel branch by @PauBadiaM in #17
• Added the bulk liana analysis at the end of bulk notebook by @adugourd in #26
• association between embeddings and sample metadata using simple anova model by @demian1 in #58
• updated docs by @demian1 in #60
• minor doc fix by @fedor-grigoryev in #40
• Updated benchmark to remove NAs in activities by @smuellerd in #62
• [WIP] Add lead_genes and n_genes to enrichment results by @pcm32 in #41
• removes dtype from anndata by @Intron7 in #79
• Fix formatting by @Zethson in #100
• Tutorial clustermap changed to Marsilea by @shuang1330 in #96

New Contributors

• @demian1 made their first contribution in #8
• @PauBadiaM made their first contribution in #17
• @adugourd made their first contribution in #26
• @fedor-grigoryev made their first contribution in #40
• @smuellerd made their first contribution in #62
• @pcm32 made their first contribution in #41
• @Intron7 made their first contribution in #79
• @Zethson made their first contribution in #100
• @shuang1330 made their first contribution in #96

Full Changelog: https://github.com/saezlab/decoupler-py/commits/v1.6.0