Show scatterplots with colorbar side-by-side

[grst]

Hi,

I created scatterplots for quality control and want to show them side-by-side to save some space.
I tried it using subplots, but this is what happens:

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
p1 = sc.pl.scatter(adata, x='n_counts', y='n_genes', color='n_genes', show=False, ax=ax1)
p2 = sc.pl.scatter(adata[adata.obs['n_counts'] < 10000], x='n_counts', y='n_genes', color='n_genes', show=False, ax=ax2)

The problem seems to be that pl.scatter adds an additional axes of that subplots doesn't know anything. Is there an easy fix for that?

Versions

scanpy==1.4.4 anndata==0.6.22.post1 umap==0.3.10 numpy==1.17.1 scipy==1.3.1 pandas==0.25.1 scikit-learn==0.21.3 statsmodels==0.10.1 python-igraph==0.7.1 louvain==0.6.1

Full repex

import scanpy as sc
import numpy as np
from matplotlib import pyplot as plt

adata = sc.read_h5ad("data/pbmc3k_raw.h5ad")

adata.obs['n_counts'] = adata.X.sum(axis=1)
adata.obs['n_genes'] = (adata.X != 0).sum(axis=1)

# show plots below each other (standard, works) 
p1 = sc.pl.scatter(adata, x='n_counts', y='n_genes', color='n_genes', show=False)
p2 = sc.pl.scatter(adata[adata.obs['n_counts'] < 10000], x='n_counts', y='n_genes', color='n_genes', show=False)

# try to show plots side-by-side (looks weird) 
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
p1 = sc.pl.scatter(adata, x='n_counts', y='n_genes', color='n_genes', show=False, ax=ax1)
p2 = sc.pl.scatter(adata[adata.obs['n_counts'] < 10000], x='n_counts', y='n_genes', color='n_genes', show=False, ax=ax2)

[VolkerBergen]

Will be fixed soon with merging scvelo's scatter. Running the same with scvelo.pl.scatter currently yields the following.

where the same color range and colorbar can be enforced by setting vmin and vmax.

[using artificial values for demonstration purposes]