使用purrr :: map将新列分配给data.tables列表

Question

I am trying to add a new column to all data.tables residing in a list. All data.tables in the list have identical structure.

A reprex: Let's take a simple list of 2 datatables ldt

library(data.table)
dt1 <- data.table(x=1:3,y=(1:3)^2)
dt2 <- data.table(x=5:7,y=(1:3)^2)
ldt <- list(d1=dt1,d2=dt2)

Now suppose I want to add a new column z in both the datatables with values stored in a character vector:

new_col <- c("DOC1","DOC2")

So I want to attach DOC1 to dt1$z and DOC2 to dt2$z

I tried the following syntax of purrr::map

purrr::map2(.x = ldt,.y = doc,.f = ~ ldt$z = doc)

But it gives the following error:

Error: unexpected '=' in "map2(.x = ldt,.y = doc,.f = ~ ldt$z ="

I tried :

 map2(.x = ldt,.y = doc,.f = function(x,y) x$z <- y)

But this results in a list output with just this:

$d1
[1] "DOC1"

$d2
[1] "DOC2"

What am I doing wrong?

Also if someone may please direct me to a purrr tutorial that takes up some different examples than the straighforward ones found here, it would be great.

https://stackoverflow.com/a/48144294/1972786 https://purrr.tidyverse.org/reference/map2.html#arguments

Answer 1

You need to use kinda unqouting. (Sorry, I'm not familiar with data.table at all, but in tidy evaluation this term sounds like this). But that's how you can solve it:

What we have...

library(data.table)
library(tidyverse)

dt1 <- data.table(x=1:3,y=1:3^2)
dt2 <- data.table(x=5:7,y=1:3^2)
ldt <- list(d1=dt1,d2=dt2)

new_col <- c("DOC1","DOC2")

Now you iterate over list of data.tables ( .x ) and values for new variable z ( .y ). data.table use non-standard evaluation inside [] , so to create new variable named with character vector "z" you have to use := . See the code:

map2(ldt, new_col, ~ .x[, "z" := .y])
#> $d1
#>    x y    z
#> 1: 1 1 DOC1
#> 2: 2 2 DOC1
#> 3: 3 3 DOC1
#> 4: 1 4 DOC1
#> 5: 2 5 DOC1
#> 6: 3 6 DOC1
#> 7: 1 7 DOC1
#> 8: 2 8 DOC1
#> 9: 3 9 DOC1
#> 
#> $d2
#>    x y    z
#> 1: 5 1 DOC2
#> 2: 6 2 DOC2
#> 3: 7 3 DOC2
#> 4: 5 4 DOC2
#> 5: 6 5 DOC2
#> 6: 7 6 DOC2
#> 7: 5 7 DOC2
#> 8: 6 8 DOC2
#> 9: 7 9 DOC2