When creating heatmaps, matrices, and various graphs using large datasets with seaborn, scanpy.set_figure_params() provides helpful default configurations for the plot. However, one default, caused by matplotlib.rcParams["axes.grid"] = True, provides no benefit to the vast majority of plots, making them unreadable and confusing (especially with very large plots). Below is a simple example; evidently, the distortion increases with graph complexity.
Images
Graph with default grid:
Graph without default grid:
Conclusion
This change simply modifies matplotlib.rcParams["axes.grid"] from True to False, not affecting any other components of the module. We believe this adjusted default would benefit the majority of graphs compared to the original. If one would like to use the grid lines for any reason (which we believe is a niche use case), they may run matplotlib.pyplot.rcParams["axes.grid"] = True.
Description
When creating heatmaps, matrices, and various graphs using large datasets with seaborn,
scanpy.set_figure_params()
provides helpful default configurations for the plot. However, one default, caused bymatplotlib.rcParams["axes.grid"] = True
, provides no benefit to the vast majority of plots, making them unreadable and confusing (especially with very large plots). Below is a simple example; evidently, the distortion increases with graph complexity.Images
Graph with default grid:
Graph without default grid:
Conclusion
This change simply modifies
matplotlib.rcParams["axes.grid"]
fromTrue
toFalse
, not affecting any other components of the module. We believe this adjusted default would benefit the majority of graphs compared to the original. If one would like to use the grid lines for any reason (which we believe is a niche use case), they may runmatplotlib.pyplot.rcParams["axes.grid"] = True
.