How can I separate a string into multiple rows in the same database in R?-CodePudding

I have a problem. This is my initial example data base:

nam <- c("Marco", "Clara")
code <- c("The liquidations code for Marco are: 51-BMR05, 74-VAD08, 176-VNF09.",
          "The liquidations code for Clara are: 88-BMR05, 90-VAD08, 152-VNF09.")

df <- data.frame(name,code)

That look like this:

So what I want is for the codes after the double dot to be separated and be a record with the same name. That is to say that the database is transformed and finished in this way:

I need to know if there is any way in R that can help me to facilitate and speed up this work. I did the examples in excel. In advance, I thank everyone for their help.

CodePudding user response：

Here is a tidy solution:

library(tidyverse)

df %>% 
  # Remove text and trailing dot
  mutate(
    code = stringr::str_remove(
      string = code, 
      pattern = "The liquidations code for .* are: "
    ),
    code = stringr::str_remove(
      string = code, 
      pattern = "\\.$"
    )
  ) %>%
  # Split the codes (results in list column)
  mutate(code = stringr::str_split(code, ", ")) %>% 
  # Turn list column into new rows
  unnest(code)


#> # A tibble: 6 × 2
#>   name  code     
#>   <chr> <chr>    
#> 1 Marco 51-BMR05 
#> 2 Marco 74-VAD08 
#> 3 Marco 176-VNF09
#> 4 Clara 88-BMR05 
#> 5 Clara 90-VAD08 
#> 6 Clara 152-VNF09

^{Created on 2022-03-28 by the reprex package (v2.0.1)}

Data

Same code as posted by OP, but fixed nam to name:

name <- c("Marco", "Clara")
code <- c("The liquidations code for Marco are: 51-BMR05, 74-VAD08, 176-VNF09.",
          "The liquidations code for Clara are: 88-BMR05, 90-VAD08, 152-VNF09.")
df <- data.frame(name,code)

CodePudding user response：

We could use str_extract_all to extract all the codes in list and then unnest

library(dplyr)
library(stringr)
library(tidyr)
df %>% 
  mutate(code = str_extract_all(code, "\\d -[A-Z0-9] ")) %>% 
  unnest(code)

-output

# A tibble: 6 × 2
  name  code     
  <chr> <chr>    
1 Marco 51-BMR05 
2 Marco 74-VAD08 
3 Marco 176-VNF09
4 Clara 88-BMR05 
5 Clara 90-VAD08 
6 Clara 152-VNF09

CodePudding user response：

I accomplished the output with this:

df1 <- separate_rows(df, code, sep=', ', convert = TRUE)
df1$liquidation_code <- gsub('The liquidations code for Marco are: |The liquidations code for Clara are: |\\.','',df1$code)
df1 <- df1[ , !(colnames(df1) %in% c('code'))]

df1

CodePudding user response：

Something like this should help

library(tidyverse)
#> Warning: package 'tidyr' was built under R version 4.1.3
#> Warning: package 'readr' was built under R version 4.1.3
#> Warning: package 'dplyr' was built under R version 4.1.3

name <- c("Marco", "Clara")
code <- c("The liquidations code for Marco are: 51-BMR05, 74-VAD08, 176-VNF09.",
          "The liquidations code for Clara are: 88-BMR05, 90-VAD08, 152-VNF09.")

df_example <- data.frame(name,code)

df_example |> 
  mutate(codes = str_extract(code,pattern = "(?<=:). ") |> 
           str_split(',')) |> 
  unnest(codes) |> 
  mutate(codes = codes |> str_squish() |> str_remove('\\.'))
#> # A tibble: 6 x 3
#>   name  code                                                               codes
#>   <chr> <chr>                                                              <chr>
#> 1 Marco The liquidations code for Marco are: 51-BMR05, 74-VAD08, 176-VNF0~ 51-B~
#> 2 Marco The liquidations code for Marco are: 51-BMR05, 74-VAD08, 176-VNF0~ 74-V~
#> 3 Marco The liquidations code for Marco are: 51-BMR05, 74-VAD08, 176-VNF0~ 176-~
#> 4 Clara The liquidations code for Clara are: 88-BMR05, 90-VAD08, 152-VNF0~ 88-B~
#> 5 Clara The liquidations code for Clara are: 88-BMR05, 90-VAD08, 152-VNF0~ 90-V~
#> 6 Clara The liquidations code for Clara are: 88-BMR05, 90-VAD08, 152-VNF0~ 152-~

^{Created on 2022-03-28 by the reprex package (v2.0.1)}

CodePudding user response：

Using regex from stringr

#build your df
name <- c("Marco", "Clara")
code <- c("The liquidations code for Marco are: 51-BMR05, 74-VAD08, 176-VNF09.",
          "The liquidations code for Clara are: 88-BMR05, 90-VAD08, 152-VNF09.")
df <- data.frame(name,code)

# Processing
codes = stringr::str_extract_all(df$code, "([:alnum:] -[:alnum:] )")
names(codes) <- c(df$name)
liquidation_code = unlist(codes)
fix_names = substr(names(liquidation_code),1,nchar(names(liquidation_code))-1)
fix_df = data.frame(names = fix_names,liquidation_code) 


       names liquidation_code
Marco1 Marco         51-BMR05
Marco2 Marco         74-VAD08
Marco3 Marco        176-VNF09
Clara1 Clara         88-BMR05
Clara2 Clara         90-VAD08
Clara3 Clara        152-VNF09