Include dplyr function rename in tidylog package

[damianooldoni]

While using this very useful package, I find it would be nice to get information also about column renaming while working with dplyr::rename() in pipes.

[elbersb]

This is a bit tricky because it’s impossible to tell how the columns were renamed from just inspecting the data frame before and after. This means one has to gather this information from the arguments, which is error-prone. Any ideas would be appreciated.

[damianooldoni]

This is a bit tricky because it’s impossible to tell how the columns were renamed from just inspecting the data frame before and after.

Maybe by comparing column names by commando colnames(df) before and after?

> a <- data.frame(col1 = c(1,2), col2 = c(3,4))
> name_cols_before <- colnames(a)
> name_cols_before
[1] "col1" "col2"
> a <- dplyr::rename(a, newcol1 = col1)
> name_cols_after <- colnames(a)
> name_cols_after 
[1] "newcol1" "col2"
> paste(
  "Column", 
  name_cols_before[name_cols_after != name_cols_before], 
  "renamed as", 
  name_cols_after[name_cols_after != name_cols_before]
)
[1] "Column col1 renamed as newcol1"

This is for dplyr::rename, likely not for multiple renames via rename_at or rename_if.

Or did I misunderstand your reply? Still thanks for paying attention to this issue.

[elbersb]

Right, that's the approach that's being used for all the other functions. But I always assumed that rename would reorder the columns, which would make this strategy impossible. It seems though that rename does not reorder columns. This would mean all the column positions match up, and it shouldn't be a problem to implement rename, rename_at, rename_all, etc. Could you check whether this understanding is correct? The docs for dplyr don't mention this explicitly, so maybe we can experiment a bit more.

a <- data.frame(col1 = c(1,2), col2 = c(3,4))
names(a)
#> [1] "col1" "col2"
names(dplyr::rename(a, newcol1 = col1, newcol2 = col2))
#> [1] "newcol1" "newcol2"
names(dplyr::rename(a, newcol1 = col2, newcol2 = col1))
#> [1] "newcol2" "newcol1"
names(dplyr::rename(a, newcol2 = col2, newcol1 = col1))
#> [1] "newcol1" "newcol2"

^{Created on 2019-09-17 by the reprex package (v0.3.0)}

[damianooldoni]

Thanks @elbersb : yes, I played a little with rename_* functions as well and no changes in order of columns detected.