Corrplot.mixed & plotCI about corrplot HOT 5 CLOSED

Could you provide a reproducible example ?
What are the objects M, L and U ?
(I assume M <- cor(mtcars))

from corrplot.

Here is the example (using corrplot_0.73)

M <- cor(mtcars)
L <- M-0.01
U <- M+0.01
corrplot::corrplot.mixed(M,lower='number',upper='circle',low=L,upp=U,plotCI='rect')

This results in:

Error in corrplot(corr, add = TRUE, type = "lower", method = lower, diag = (diag == : method shoud be circle or square if draw confidence interval!

from corrplot.

I fixed the error for now, but further API redesign is needed, I think.
Here is the output generated from your example code:

Please confirm whether this is the expected output.
If yes, please create a new issue and suggest the API change.

from corrplot.

This indeed works for me!

It only works for numbers though. Other cases:

corrplot::corrplot.mixed(M,lower='circle',upper='square',low=L,upp=U,plotCI='rect')
(lower boxplot, upper boxplot - expected: lower circle, upper boxplot)

corrplot::corrplot.mixed(M,lower='ellipse',upper='square',low=L,upp=U,plotCI='rect')
(lower empty, upper boxplot - expected: lower ellipse, upper boxplot)

corrplot::corrplot.mixed(M,lower='pie',upper='square',low=L,upp=U,plotCI='rect')
results in:

Error in corrplot(corr, add = TRUE, type = "lower", method = lower, diag = (diag == :
method shoud be circle or square if draw confidence interval!

I am not sure what kind of API redesign you are thinking of because I did not look into the underlying code. It is probably more helpful if you post the api change, because you know what needs to change in the api.

from corrplot.

I'm closing this for now, because it seems that I covered the use-case with my latest fix.

from corrplot.