Regarding the function 'get_metadata_association'

[saezlab]

Regarding the function 'get_metadata_association'

[demian1]

Hi @Phil-Momoka
This function was added just last week and is therefore only available through GitHub for now.
You can update to the version on GitHub with pip install git+https://github.com/saezlab/decoupler-py.git
Be aware that the association is made with ANOVA using the statsmodels package that you need to install as well.

[Phil-Momoka]

Hi @demian1
Thank you for your quick response! I will try using the newest version.

[ssun1116]

Hello, I'm also interested in using the get_metadata_associations function. It has been working well for me so far. However, I encountered a TypeError when attempting to use the dc.plot_associations code.
I would greatly appreciate it if you could help me resolve this issue.
Thank you in advance!

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[14], line 2
      1 plt.figure(figsize=(7,10))
----> 2 ax, legend_axes = dc.plot_associations(pp_pdata,
      3                                        uns_key='pca_anova', #summary statistics from the anova tests
      4                                        obsm_key='X_pca', #where the PCs are stored
      5                                        stat_col='p_adj', #which summary statistic to plot
      6                                        obs_annotation_cols = ['condition', 'cluster_original'],
      7                                        titles=['Adjusted p-values from ANOVA', 'Principle component scores']) # which sample annotations to plot
      8 plt.show()

File ~/.conda/envs/BrainAtlas/lib/python3.8/site-packages/decoupler/plotting.py:1590, in plot_associations(data, uns_key, associations, cols, obs_annotation_cols, obsm_key, use_X, layer, stat_col, titles, scores_kwargs, stats_kwargs)
   1587 cm_stats = pch.ClusterMapPlotter(data=stats, plot=False, **stats_defaults)
   1589 #combine clustermaps vertically
-> 1590 ax, legend_axes = pch.composite(cmlist = [cm_stats, cm_scores], main = 1, axis = 0, height_ratios=[1,3], row_gap=7, legend_gap=15)
   1592 #add titles
   1593 if titles is not None:

TypeError: composite() got an unexpected keyword argument 'height_ratios'

[demian1]

Hi @ssun1116
In order to fix this issue, I would recommend to update to the latest version of PyComplexHeatmap (at least commit ac50aca) from GitHub
pip install git+https://github.com/DingWB/PyComplexHeatmap