Arrange by custom order in dplyr using dynamic column names

Question

I am trying to arrange a tibble based on a certain column (but setting the column based on a name dynamically set by a variable).

Below is the code I currently am trying but am getting an error. The second paste of code works (where I have hard coded in the column name of symbol which I want to be set based off of a variable instead).

library(tidyverse)

group_var <- "symbol"

date_seq <- seq(as.Date("2000-01-01"), as.Date("2009-12-31"), by = "days")
test_tbl <- tibble::tibble("date" = rep(date_seq, 3),
                           "symbol" = rep(c("test3", "test1", "test2"), each = length(date_seq)),
                           "value" = c(rnorm(length(date_seq), sd = 0.05),
                                       rnorm(length(date_seq), sd = 0.05),
                                       rnorm(length(date_seq), sd = 0.05)))

order_var <- c("test1", "test2", "test3")
test_tbl_final <- test_tbl %>%
  dplyr::arrange(factor(!!group_var, levels = order_var), date)

Below is the code that works and shows what I am trying to accomplish:

library(tidyverse)

date_seq <- seq(as.Date("2000-01-01"), as.Date("2009-12-31"), by = "days")
test_tbl <- tibble::tibble("date" = rep(date_seq, 3),
                           "symbol" = rep(c("test3", "test1", "test2"), each = length(date_seq)),
                           "value" = c(rnorm(length(date_seq), sd = 0.05),
                                       rnorm(length(date_seq), sd = 0.05),
                                       rnorm(length(date_seq), sd = 0.05)))

order_var <- c("test1", "test2", "test3")
test_tbl_final <- test_tbl %>%
  dplyr::arrange(factor(symbol, levels = order_var), date)

IceCreamToucan · Accepted Answer · 2018-02-27 18:49:48Z

7

You can also use as.symbol from base R

test_tbl_final <- test_tbl %>%
  dplyr::arrange(factor(!!as.symbol(group_var), levels = order_var), date)

answered Feb 27, 2018 at 18:49

IceCreamToucan

28.8k2 gold badges27 silver badges48 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Trevor Nederlof Over a year ago

Thank you, love the base R function mention.

akuiper · Accepted Answer · 2018-02-27 18:49:09Z

3

You need rlang:sym to convert group_var from character symbol to a symbol symbol and then use !! to evaluate the symbol as a column object:

test_tbl %>% 
    arrange(factor(!!rlang::sym(group_var), levels = order_var), date)

# A tibble: 10,959 x 3
#         date symbol         value
#       <date>  <chr>         <dbl>
# 1 2000-01-01  test1  0.0519143671
# 2 2000-01-02  test1 -0.0464782439
# 3 2000-01-03  test1 -0.0295441613
# ...

answered Feb 27, 2018 at 18:49

akuiper

216k33 gold badges362 silver badges379 bronze badges

Collectives™ on Stack Overflow

Arrange by custom order in dplyr using dynamic column names

2 Answers 2

1 Comment

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related