Alternative pivoting method for easily pivoting based on name pattern
Source:R/wide2long.R
wide2long.RdThis function requires and assumes a systematic naming of variables. For now only supports one level pivoting. Adding more levels would require an added "ignore" string pattern or similarly. Example 2.
Usage
wide2long(
data,
pattern,
type = c("prefix", "infix", "suffix"),
id.col = 1,
instance.name = "instance"
)Examples
data.frame(
1:20, sample(70:80, 20, TRUE),
sample(70:100, 20, TRUE),
sample(70:100, 20, TRUE),
sample(170:200, 20, TRUE)
) |>
setNames(c("id", "age", "weight_0", "weight_1", "height_1")) |>
wide2long(pattern = c("_0", "_1"), type = "suffix")
#> id age instance weight height
#> 1 1 73 0 93 NA
#> 2 1 NA 1 80 189
#> 3 2 76 0 71 NA
#> 4 2 NA 1 78 185
#> 5 3 73 0 71 NA
#> 6 3 NA 1 89 196
#> 7 4 76 0 81 NA
#> 8 4 NA 1 73 184
#> 9 5 70 0 98 NA
#> 10 5 NA 1 85 183
#> 11 6 70 0 90 NA
#> 12 6 NA 1 99 183
#> 13 7 77 0 74 NA
#> 14 7 NA 1 77 186
#> 15 8 80 0 81 NA
#> 16 8 NA 1 72 200
#> 17 9 72 0 95 NA
#> 18 9 NA 1 90 173
#> 19 10 79 0 72 NA
#> 20 10 NA 1 92 184
#> 21 11 78 0 90 NA
#> 22 11 NA 1 90 194
#> 23 12 72 0 83 NA
#> 24 12 NA 1 79 194
#> 25 13 76 0 80 NA
#> 26 13 NA 1 75 172
#> 27 14 76 0 100 NA
#> 28 14 NA 1 76 173
#> 29 15 74 0 83 NA
#> 30 15 NA 1 96 181
#> 31 16 75 0 80 NA
#> 32 16 NA 1 79 199
#> 33 17 77 0 71 NA
#> 34 17 NA 1 95 178
#> 35 18 76 0 89 NA
#> 36 18 NA 1 99 175
#> 37 19 75 0 78 NA
#> 38 19 NA 1 88 175
#> 39 20 79 0 83 NA
#> 40 20 NA 1 98 178
data.frame(
1:20, sample(70:80, 20, TRUE),
sample(70:100, 20, TRUE),
sample(70:100, 20, TRUE),
sample(170:200, 20, TRUE)
) |>
setNames(c("id", "age", "weight_0", "weight_a_1", "height_b_1")) |>
wide2long(pattern = c("_0", "_1"), type = "suffix")
#> id age instance weight weight_a height_b
#> 1 1 78 0 74 NA NA
#> 2 1 NA 1 NA 87 171
#> 3 2 73 0 78 NA NA
#> 4 2 NA 1 NA 99 186
#> 5 3 77 0 71 NA NA
#> 6 3 NA 1 NA 92 196
#> 7 4 75 0 80 NA NA
#> 8 4 NA 1 NA 71 174
#> 9 5 80 0 71 NA NA
#> 10 5 NA 1 NA 88 173
#> 11 6 77 0 74 NA NA
#> 12 6 NA 1 NA 79 195
#> 13 7 72 0 90 NA NA
#> 14 7 NA 1 NA 86 200
#> 15 8 76 0 93 NA NA
#> 16 8 NA 1 NA 99 175
#> 17 9 79 0 99 NA NA
#> 18 9 NA 1 NA 95 192
#> 19 10 79 0 81 NA NA
#> 20 10 NA 1 NA 97 179
#> 21 11 76 0 81 NA NA
#> 22 11 NA 1 NA 74 175
#> 23 12 75 0 90 NA NA
#> 24 12 NA 1 NA 93 186
#> 25 13 77 0 94 NA NA
#> 26 13 NA 1 NA 71 179
#> 27 14 79 0 76 NA NA
#> 28 14 NA 1 NA 92 200
#> 29 15 76 0 99 NA NA
#> 30 15 NA 1 NA 96 184
#> 31 16 72 0 78 NA NA
#> 32 16 NA 1 NA 92 196
#> 33 17 73 0 84 NA NA
#> 34 17 NA 1 NA 80 180
#> 35 18 76 0 75 NA NA
#> 36 18 NA 1 NA 95 199
#> 37 19 77 0 84 NA NA
#> 38 19 NA 1 NA 79 170
#> 39 20 80 0 93 NA NA
#> 40 20 NA 1 NA 97 190
# Optional filling of missing values by last observation carried forward
# Needed for mmrm analyses
long_missings |>
# Fills record ID assuming none are missing
tidyr::fill(record_id) |>
# Grouping by ID for the last step
dplyr::group_by(record_id) |>
# Filling missing data by ID
tidyr::fill(names(long_missings)[!names(long_missings) %in% new_names]) |>
# Remove grouping
dplyr::ungroup()
#> Error: object 'long_missings' not found