Suggest reversing the color key to be red to blue

[kwstat]

Corrgrams were popularized by Friendly in this paper: www.datavis.ca/papers/corrgram.pdf

He used a red (negative) to blue (positive) color palette, which has since become a sort of standard for corrgram colors.

I think your reverse color scheme will confuse many people.

[kassambara]

Thank you for this suggestion. I will update the package to use the standard color scheme. In the meantime, user can change color as suggested at: http://www.sthda.com/english/wiki/ggcorrplot-visualization-of-a-correlation-matrix-using-ggplot2

[kwstat]

Thanks.

[jabbate]

Hi all. I randomly fell on this exchange, as I wanted to reverse the color scheme - because WHY on earth would anyone think red (hot) intuitively represents negative and blue (cold) represents positive values? I suggest we all go back to something that makes sense. ;)

[kwstat]

The intuitive interpretation is that negative correlation is bad (hence red) while positive correlation is good (some would use green, but blue is colorblind safe). (Think about levelplots with bad/good areas and it might make more sense). Any other interpretation is suspect. :-)

[jeffgoblue]

I'm not sure about the intuitive "negative correlation is bad, positive correlation is blue" approach... the sign simply implies directionality of the correlation (i.e. proportional vs inversely proportional)). In sales, price and volume are negatively correlated, but that's not a bad thing. Personally, my intuition harkens to red being "hotter" and thus "higher". But we all have our own lenses for viewing things.

My question is, what is the simplest way to reverse the default scale on ggcorrplot so it can be to my liking?