Pandas-profiling: Add functionality to modify plots (and their look)

[jankaWIS]

Missing functionality
I have searched pandas profiling documentation, the advanced usage webpage and also the plotting but I could not find how to customise plots that it generates. Eg. now it shows correlation matrices with a colorbar having blue as 1 and red as -1. It would be great if one could play with the plots (eg pass some arguments to the underlying matplotlib) to change the colours, the palette, the labels/legend, ...

Proposed feature
Be able to add eg cmap argument or color argument to the plotting function to change the look of the graphs. The same way as one can skip some correlations using the advanced options for usage.

Thanks, your work is really great!

[sbrugman]

It's already possible to change the cmap, however it's not directly documented (you can find the related parameters in the config file). Feel free to extend the docs with the relevant settings in a PR.

[jankaWIS]

Perfect, thanks a lot. How detailed should it be? Do you have some guidelines on how to write your docs? Or is it enough to load some dataset and show on some examples how to make it, eg. as seaborn does.

[sbrugman]

The addtions should be consistent with the existing docs and helpful to readers. You can find the relevant page here: https://github.com/pandas-profiling/pandas-profiling/blob/develop/docsrc/source/pages/advanced_usage.rst. Feel free to add in examples or anything you feel is useful to the reader. Much appreciated!