How to order a DataTable using a hidden column - r

Question

I am rather new to R and I am trying to prepare an interactive data table using the DT package. My data contains numeric values, but some of these values are preceded by < or > sign. What I want is for my data table is to allow interactive sorting on the numeric values, regardless of whether there is a < or > sign in front of it. So for example >10, <5, 9, >8 should sort to <5, >8, 9, >10.
My initial approach for this was to duplicate the column containing the numeric values with < and > signs, to remove the < and > signs from this duplicate column, and to convert this data to numeric values to obtain a column with only the numeric values. What I then would like is to be able to order the data in the table on these numeric values, but I want to be able to do this when clicking the ordening button of the column containing the numeric values with the < and > signs. Therefore, I want to hide the column containing only the numeric values (since I do not want this column to be present in the table), but I want to somehow link the ordining function of the original column to this hidden column.
Here are some example data and a script in which I have already duplicated the column (b to c), removed the < and > signs, and converted it to numeric values to obtain the column c, which I have then hidden:
library(DT)
df <- data.frame(a=1:5, b=c('10','5.0','2.0','< 1.0','> 20'), c=c(10,5,2,1,20))
DT <- DT::datatable(df,
options = list(columnDefs =
list(list(visible=FALSE,
targets=3))))
DT
I have not been able to find a way to sort the data in the table on this hidden column c by using the sorting button of column b.
I have found that this should be possible in JavaScript: jQuery DataTables - Ordering dates by hidden column
However, I am not able to figure out how to do the same in R, either by using a suitable function in R, or by providing it in JavaScript using the JS() function.
Could anyone help me with this problem?

Answer 1

Here is a solution using render:
library(DT)
render <- c(
"function(data, type, row){",
" if(type === 'sort'){",
" return parseFloat(data.match(/\\d+\\.?\\d+/)[0]);",
" }else{",
" return data;",
" }",
"}"
)
df <- data.frame(
a = 1:5,
b = c('10','5.0','2.0','< 1.0','> 20')
)
DT <- datatable(df,
options = list(
columnDefs = list(
list(render = JS(render), type = "num", targets = 2)
)
)
)
DT
This solution does not require a hidden column.

Answer 2

Here's a way to do it. To get the "sorting key" use order.
library(DT)
# df <- data.frame(a=1:5, b=c('10','5.0','2.0','< 1.0','> 20'), c=c(10,5,2,1,20))
df <- data.frame(a = 1:5, b = c('10', '5.0', '2.0', '< 1.0', '> 20'))
df
#ONE APPROACH
df$c <-
stringr::str_replace(string = df$b,
pattern = "[<>]",
replacement = "") %>%
as.numeric()
#ANOTHER APPROACH
df$c <- gsub("[<>]", "", df$b) %>% as.numeric()
DT::datatable(df[order(df$c), -3], rownames = FALSE)

Answer 3

library(DT)
df <- data.frame(a=1:5, b=c('10','5.0','2.0','< 1.0','> 20'), c=c(10,5,2,1,20))
DT <- DT::datatable(df,
options = list(columnDefs =
list(list(visible=FALSE, targets=3),
list(orderData=3, targets=2)
)))
DT
Note: This answer is based on this one here, but DT now uses R indexing instead of JS indexing.