`correlation()` error when `data2` is a `grouped_df` object about correlation HOT 2 CLOSED

Yeah, this is a rather tricky and challenging issue, not straightforward to address because of the way R works (as it dispatches the methods on the first argument). Not sure if we can fix that 🤔

Not even sure of what the output would look like? each correlation for each combination of groups?

from correlation.

But I found #47 that there was still a problem:

correlation(dplyr::group_by(iris, Species),
                   dplyr::group_by(iris, Species))

## Error in `$<-.data.frame`(`*tmp*`, "Group", value = "setosa") : 
## replacement has 1 row, data has 0

Let me look at the logic of correlation_grouped_df().

from correlation.