Is there an R function to move data after a certain string? - r

Question

I am trying to move data to a new column after certain points in the data. My data is spread across multiple data frames that only have some elements in common, so I would like to be able to make a loop to clean the data sets. I am looking for a function that after the first time there is certain text, for example "Total", in a row all the data below that moves to a new columns.
first
second
third
One
1
One
1
Total
2
Two
2
Two
2
Total
2
I want my data to look similar to this below, but due to the variability of the data I am having trouble finding a solution that can be reproduced easily.
left
center
right
fourth
One
1
Two
2
One
1
Two
2
Total
1
Total
2

Answer 1

Personal opinion cbinding data on wider side will be too cumbersome, if the rows are too much. Still you can divide the data into separate groups like this
df <- read.table(text = "first second
One 1
One 1
Total 2
Two 2
Two 2
Total 2", header = T)
df$dummy = rev(cumsum(rev(df$first == "Total")))
df
> df
first second dummy
1 One 1 2
2 One 1 2
3 Total 2 2
4 Two 2 1
5 Two 2 1
6 Total 2 1
You may notice that your data is divided into two groups. You may still cbind() or bind_cols() if you want, easily
df %>% group_split(d = rev(cumsum(rev(first == "Total")))) %>% bind_cols()
# A tibble: 3 x 6
first...1 second...2 d...3 first...4 second...5 d...6
<chr> <int> <int> <chr> <int> <int>
1 Two 2 1 One 1 2
2 Two 2 1 One 1 2
3 Total 2 1 Total 2 2

Answer 2

Here is another try
library(dplyr, warn.conflicts = FALSE)
library(tidyr)
library(purrr)
data <- structure(list(first = c("One", "One", "Total", "Two", "Two",
"Total"), second = c(1L, 1L, 2L, 2L, 2L, 2L)), row.names = c(NA,
-6L), class = "data.frame")
new_data <- data %>%
# create group using first == "Total"
mutate(total_group = cumsum(first == "Total")) %>%
mutate(total_group = if_else(first == "Total", total_group - 1L, total_group)) %>%
# split df into multiple df and bind cols
group_split(total_group, .keep = FALSE) %>%
bind_cols()
#> New names:
#> * first -> first...1
#> * second -> second...2
#> * first -> first...3
#> * second -> second...4
new_data
#> # A tibble: 3 x 4
#> first...1 second...2 first...3 second...4
#> <chr> <int> <chr> <int>
#> 1 One 1 Two 2
#> 2 One 1 Two 2
#> 3 Total 2 Total 2
# if you only have two group this could work - otherwise need some more
# work on the approach. Hope this provide you enough hint to develop further
names(new_data) <- c("left", "center", "right", "fourth")
new_data
#> # A tibble: 3 x 4
#> left center right fourth
#> <chr> <int> <chr> <int>
#> 1 One 1 Two 2
#> 2 One 1 Two 2
#> 3 Total 2 Total 2
Created on 2021-04-04 by the reprex package (v1.0.0)