New plot function to draw the relations between 2 categorical adata.obs variables?

What kind of feature would you like to request?

New plotting function: A kind of plot you would like to seein sc.pl?

Please describe your wishes

Hello Scanpy,
Thank you for developing this amazing package.
I have 2 categorical variables in adata.obs

adata_sub2.obs['patient']
AAACCTGAGGTTACCT-1-C106_N_1_1_0_c1_v2    C106_N
AAACCTGGTCAGAAGC-1-C106_N_1_1_0_c1_v2    C106_N
AAACCTGGTGCTCTTC-1-C106_N_1_1_0_c1_v2    C106_N
AAACCTGTCCCATTAT-1-C106_N_1_1_0_c1_v2    C106_N
AAACCTGTCGGAGCAA-1-C106_N_1_1_0_c1_v2    C106_N
                                          ...  
TTTGCGCAGACACGAC-1-SMC25T                SMC25T
TTTGCGCCATGGAATA-1-SMC25T                SMC25T
TTTGCGCCATGTTCCC-1-SMC25T                SMC25T
TTTGGTTGTAGGGTAC-1-SMC25T                SMC25T
TTTGTCAAGAGGGATA-1-SMC25T                SMC25T
Name: patient, Length: 44245, dtype: category
Categories (39, object): ['C106_N', 'C126_N', 'C130_N', 'C133_N', ..., 'C138_T', 'C168_T', 'SMC17T', 'SMC20T']

adata_sub2.obs['type']
AAACCTGAGGTTACCT-1-C106_N_1_1_0_c1_v2    Normal
AAACCTGGTCAGAAGC-1-C106_N_1_1_0_c1_v2    Normal
AAACCTGGTGCTCTTC-1-C106_N_1_1_0_c1_v2    Normal
AAACCTGTCCCATTAT-1-C106_N_1_1_0_c1_v2    Normal
AAACCTGTCGGAGCAA-1-C106_N_1_1_0_c1_v2    Normal
                                          ...  
TTTGCGCAGACACGAC-1-SMC25T                 NMCAD
TTTGCGCCATGGAATA-1-SMC25T                 NMCAD
TTTGCGCCATGTTCCC-1-SMC25T                 NMCAD
TTTGGTTGTAGGGTAC-1-SMC25T                 NMCAD
TTTGTCAAGAGGGATA-1-SMC25T                 NMCAD
Name: type, Length: 44245, dtype: category
Categories (3, object): ['Normal', 'NMCAD', 'MCAD']

I know Scanpy has sc.pl.matrixplot(), and we can make matrix plot for gene set score, like

sc.pl.matrixplot(adata_sub2, var_names=['Normal_signature'], groupby='patient', use_raw=True, dendrogram=False, cmap='inferno', standard_scale='var', swap_axes=True, save='27.pdf')

Could you please let me know whether Scanpy can make a matrixplot-like plot showing the 'type' grouped by 'patient'? like:

sc.pl.matrixplot(adata_sub2, var_names=['type'], groupby='patient', palette={'Normal': 'blue', 'NMCAD': 'green', 'MCAD': 'red'}, dendrogram=False, swap_axes=True, save='27.pdf')

Thank you!
Best,
Yuanjian

Start with the existing sc.pl.matrixplot entry point and the examples using adata.obs['patient'] and adata.obs['type']. Clarify the expected matrixplot-like behavior for two categorical variables, including palette handling and axis grouping; done means the requested relationship can be plotted through the sc.pl interface.