Conditional formatting - wrongly applying format to top row

[awechsler1]

Hi David,

I emailed you about this earlier, but here's a little more explanation of what exactly is going on:

Let's say my initial data set consists of the following columns: group, id, count, rank, coefficient, probability. I am then making a huxtable with distinct ids as the rows and with the columns labeled as the "effect" for each group - within each table cell, I have a + or - depending on the coefficient. I then use the following code to loop through each group name within the huxtable and format the table accordingly: huxtable_name -> huxtable_name %>% set_text_color(where(as.numeric(unlist(huxtable_name[,paste(group_name, "Count", sep = "_")])) < 20), which(colnames(huxtable_name) == paste(group_name, "Effect", sep = "_")), "gray") %>% set_text_color(where(as.numeric(unlist(huxtable_name[,paste(group_name, "Rank", sep = "_")])) <= 5.), which(colnames(huxtable_name) == paste(group_name, "Effect", sep = "_")), "red") %>% set_background_color(where(as.numeric(unlist(huxtable_name[,paste(group_name, "Prob", sep = "_")])) < .05), which(colnames(huxtable_name) == paste(group_name, "Effect", sep = "_")), "light blue")

Unfortunately, what ends up happening is that in the very top row sometimes, the + or - sign for various groups will turn red even if that particular id is not in the top 5 rank...but if there is an id further down that same column that does have a top 5 rank. Everything else works perfectly, and the code itself doesn't seem wrong; when checking the boolean statement for that top row, the TRUE/FALSE values align with whether or not the rank is less than 5, but the formatting seems to happen anyway.

Let me know if any of this isn't clear!

[hughjonesd]

Could you provide a self-contained example that I can run? Provide toy data with dput. My first thought is that you are trying to use where to select rows, but where returns a matrix and should be used only in the 3-argument form of set_* calls. If so, this is a bug in the code and not in huxtable. But I may be wrong.

[awechsler1]

This is some sample data - the code seems to work perfectly everywhere else. huxtable_example.txt

[hughjonesd]

Please could you provide a minimal, self-contained example? I tried to run the code in the text file and got:

Error: unexpected '<' in:
"        "Group_7_Effect"), class = c("data.table", "data.frame"), row.names = c(NA, 
          -48L), .internal.selfref = <"

In particular, I am more likely to be able to help if it is clear that the error can be reproduced in a simple example that should obviously work (e.g. not using where in the 4-argument form of a set_ call).

For reference on how to provide a MCVE, see https://stackoverflow.com/help/mcve and https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example

[hughjonesd]

Closing. Please reopen if you still feel there is a bug!