ggcorr() suggestion, add option to provide custom labels for the variables

[go4ino]

For the function ggcorr() I think adding an option to change the variable labels with an input field in the function call would be great addition.

I know you can kind of accomplish this by renaming the columns in the original dataframe, but that runs into issues if you'd like more complex names like Mug Moment Aggregate %age, and then you will want to change the column names back for other functions.

For example with the current ggcorr

library(tidyverse)
library(GGally)
data("economics")
ggcorr(economics, label = TRUE)

And the new input option could look something along the lines of

library(tidyverse)
library(GGally)
data("economics")

ggcorr(economics, label = TRUE, var_labels = c('pers_consum_expen', 'population', 'ps_rt', 'bigChungus', 'unemployed')  )

[sidocleux]

Hello. I am new to the package and together with @lauredrs and
under the supervision of @katossky and @JMeyer31 we are willing to solve this
issue, as part of an assignment at the ENSAI school of statistical engineering
(France). We have read the opensource guide of git and we will try to be as
autonomous as possible but we may ask for guidance from now and then. We
have a solid background in statistics and a reasonable one in computer science.
If any maintainer thinks this issue is too ambitious for a beginner, please tell
us so and - if you have time - guide us to a project / issue were our help may
be more useful.

[katossky]

I am not completely convinced it would be easier to use than renaming the variables, but that's a good exercice for @sidocleux and @lauredres. Have you localised where in the code of ggcorr the variable labelling do happen ? Remember to also change the documentation.

[larmarange]

Why not use variable labels, as introduced in many packages? cf. labelled::set_variable_labels(). They are already in the RStudio viewer and will be used by default in the next version of ggplot2, etc.

To be consistent, if custom variable labels are passed to the function, it would be better to have a named vector.

In that case, you could consider for a variable:

• custom label passed directly to the function;
• if not current variable label attached to the vector in the data
• if not, name of the column

[larmarange]

see tidyverse/ggplot2#6077 for introduction in next version of ggplot2