dplyr :: select（num_range（））当数字在列名的中间时

Question

In the data below, there are column names for units (1 - 8). Each unit has a column for the score and for the percent. Is there a way to use dplyr::select() with the num_range() helper to select, say, only units 1-3 for the scores? I can get it if I drop the suffix (so it's just unit_1 instead of unit_1_score), but otherwise my attempts have been unsuccessful. I've tried dplyr::select(d, num_range("unit_", 1:3, "_score")) but that doesn't seem to work. Any help would be appreciated.

d <- readr::read_csv("https://data.jacksonms.gov/api/views/97iy-g8hk/rows.csv")
d <- janitor::clean_names(d)
names(d)

 [1] "test_year"             "test_type"             "test_site"             "student_id"           
 [5] "pre_test_score"        "pre_test_percent"      "post_test_score"       "post_test_percent"    
 [9] "percentage_change"     "unit_1_score"          "unit_1_percent"        "unit_2_score"         
 [13] "unit_2_percent"        "unit_3_score"          "unit_3_percent"        "unit_4_score"         
 [17] "unit_4_percent"        "unit_5_6_score"        "unit_5_6_percent"      "unit_7_score"         
 [21] "unit_7_percent"        "unit_8_score"          "unit_8_percent"        "total_score"          
 [25] "total_percent_correct"

Answer 1

我们可以使用dplyr::matches()选择具有正则表达式范围的列：

select(d, matches("unit_[1-3]_score"))

Answer 2

I hope this answer is not perceived as off-topic; I am assuming you would be happy with a valid response even if it does not use dplyr .

You can easily select certain columns in a data.frame using regular expressions. To select only units 1-3, for example, try: d[, grep(x = colnames(d), pattern = "^unit_[1-3]{1}_.*$)"] This will select the columns in d that have column names starting with "unit_" followed by 1, 2, or 3 (only one time), and then zero or more of anything afterwards.

Answer 3

Notwithstanding that the 5_6 column is going to be tricky (who thought that was a good idea!?), you might find the new tidyeval concepts useful for this. The syms function in the rlang package and the new !!! expansion method work together to solve this kind of problem:

dplyr::select(d, !!!rlang::syms(paste0("unit_", 1:3, "_score")))
#> # A tibble: 48 x 3
#>    unit_1_score unit_2_score unit_3_score
#>           <int>        <int>        <int>
#>  1            3            4            6
#>  2            5            5            6
#>  3            4            4            6
#>  4            4            4            6
#>  5            2            5            6
#>  6            5            5            7
#>  7            5            5            6
#>  8            4            5            5
#>  9            6            4            5
#> 10            4            5            5
#> # ... with 38 more rows

Explaining exactly what this does is somewhat tricky (try reading vignette("tidy-evaluation") ) but it works, so there's that :)

Though actually, just using strings works now so maybe you don't need to bother?

dplyr::select(d, paste0("unit_", 1:3, "_score"))
#> # A tibble: 48 x 3
#>    unit_1_score unit_2_score unit_3_score
#>           <int>        <int>        <int>
#>  1            3            4            6
#>  2            5            5            6
#>  3            4            4            6
#>  4            4            4            6
#>  5            2            5            6
#>  6            5            5            7
#>  7            5            5            6
#>  8            4            5            5
#>  9            6            4            5
#> 10            4            5            5
#> # ... with 38 more rows