Skip to contents

Identify coe columns in a tibble

Source: R/utils-listcolumns.R
get_coe_cols.Rd

Detect and return the name of coefficient column(s) in a tibble. These functions look for columns with class "coe" in their class hierarchy.

Usage

get_coe_cols(df, .cols = NULL)

get_all_coe_cols(df)

Arguments

df

A data frame or tibble

.cols

Optional. Explicitly specify which column(s) to use. Can be:

  • Character vector of column names

  • Integer vector of column positions

  • Logical vector If NULL (default), automatically detects columns with class "coe".

Value

  • get_coe_cols(): Character string with single column name

  • get_all_coe_cols(): Character vector with all coe column names

Details

Detection strategy

Both functions look for columns that have "coe" anywhere in their class hierarchy. This includes:

  • Columns with class c("coe", "list")

  • Columns with class c("eft", "coe", "list")

  • Columns with class c("opoly", "coe", "list")

  • Columns with class c("npoly", "coe", "list")

  • Columns with class c("dct", "coe", "list")

  • Any other combination where "coe" is present

All morphometric methods automatically add the "coe" class to their output columns.

Differences between functions

  • get_coe_cols(): Returns a single column name. Errors if:

    • No "coe" columns found

    • Multiple "coe" columns found (user must specify .cols)

  • get_all_coe_cols(): Returns all coe column names as a vector. Errors only if:

    • No "coe" columns found

Manual specification

If automatic detection fails or you want to use a specific column, use the .cols argument in get_coe_cols():

get_coe_cols(df, .cols = "VD_coe")
get_coe_cols(df, .cols = 2)  # Second column

See also

get_coo_cols(), get_all_coo_cols(), opoly(), npoly(), dct(), eft()

Examples

# Single coe column
data(bot)
bot_eft <- bot %>% eft()
#> Using coo column: 'coo'
#> Using nb_h = 6 harmonics (default)
get_coe_cols(bot_eft)
#> [1] "coe"
# [1] "coe"

# Multiple coe columns
data(olea)
olea_poly <- olea %>% opoly()
#> Processing coo column(s): VD, VL
#> Using degree = 5 (default)
get_all_coe_cols(olea_poly)
#> [1] "VD_coe" "VL_coe"
# [1] "VD_coe" "VL_coe"

# get_coe_cols() errors with multiple
try(get_coe_cols(olea_poly))
#> Error in get_coe_cols(olea_poly) : 
#>   Multiple coe columns found: VD_coe, VL_coe. Use .cols to specify which one.
# Error: Multiple coe columns found: VD_coe, VL_coe

# But can specify explicitly
get_coe_cols(olea_poly, .cols = "VD_coe")
#> [1] "VD_coe"
# [1] "VD_coe"

# Use with inverse transforms
olea %>%
  opoly() %>%
  opoly_i(.cols = get_coe_cols(., "VD_coe"))
#> Processing coo column(s): VD, VL
#> Using degree = 5 (default)
#> # A tibble: 90 × 8
#>    id    VD        VL        var   status VD_coe  VL_coe  VD_coe_i 
#>    <chr> <cur>     <cur>     <fct> <fct>  <opoly> <opoly> <opn>    
#>  1 Ag1   (99 x 2)  (96 x 2)  Aglan cult   6 op    6 op    (120 x 2)
#>  2 Ag2   (95 x 2)  (104 x 2) Aglan cult   6 op    6 op    (120 x 2)
#>  3 Ag3   (101 x 2) (94 x 2)  Aglan cult   6 op    6 op    (120 x 2)
#>  4 Ag4   (100 x 2) (93 x 2)  Aglan cult   6 op    6 op    (120 x 2)
#>  5 Ag5   (102 x 2) (108 x 2) Aglan cult   6 op    6 op    (120 x 2)
#>  6 Ag6   (98 x 2)  (93 x 2)  Aglan cult   6 op    6 op    (120 x 2)
#>  7 Ag7   (100 x 2) (107 x 2) Aglan cult   6 op    6 op    (120 x 2)
#>  8 Ag8   (100 x 2) (93 x 2)  Aglan cult   6 op    6 op    (120 x 2)
#>  9 Ag9   (99 x 2)  (94 x 2)  Aglan cult   6 op    6 op    (120 x 2)
#> 10 Ag10  (100 x 2) (95 x 2)  Aglan cult   6 op    6 op    (120 x 2)
#> # ℹ 80 more rows