CODEWITHSUNDEEP
rsplittidyrstrsplit
D. Studer
D. Studer

Reputation: 1875

Splitting column into multiple columns by separator

I have a data.frame with a column "offence". Every offence consists of an article (Art), a paragraph (Abs) and a sub-paragraph (Ziff) :

df<-data.frame(offence=c("Art. 110 Abs. 3 StGB","Art. 10 Abs. 1 StGB", "Art. 122 SVG", "Art. 1 Ziff. 2 UWG"))

> df
               offence
1 Art. 110 Abs. 3 StGB
2  Art. 10 Abs. 1 StGB
3         Art. 122 SVG
4   Art. 1 Ziff. 2 UWG

But I need to have it in that form:

  Art Ziff Abs  Law
1 110   NA   3 StGB
2  10   NA   1 StGB
3 122   NA  NA  SVG
4   1    2  NA  UWG

What is the best way to get this result? lapply?

Thank you!

Upvotes: 1

Views: 224

Answers (2)

G. Grothendieck
G. Grothendieck

Reputation: 269441

Convert it to dcf form (i.e. keyword: value) using gsub and then read it in using read.dcf. At the end convert the matrix that read.dcf produces to a data frame and convert any number columns to numeric. No packages are used.

s <- gsub("(\\S+)[.] (\\d+)", "\\1: \\2\n", df[[1]]) # convert to keyword: value
s <- sub(" (\\D+)$", "Law: \\1\n\n", s) # handle Law column
us <- trimws(unlist(strsplit(s, "\n")))  # split into separate components
DF <- as.data.frame(read.dcf(textConnection(us)), stringsAsFactors = FALSE)
DF[] <- lapply(DF, type.convert)

giving:

  Art Abs  Law Ziff
1 110   3 StGB   NA
2  10   1 StGB   NA
3 122  NA  SVG   NA
4   1  NA  UWG    2

Upvotes: 1

acylam
acylam

Reputation: 18661

You can use str_extract from stringr:

library(stringr)
library(dplyr)

df$offence %>%
  {data.frame(Art = str_extract(., "(?<=Art[.]\\s)\\d+"),
              Ziff = str_extract(., "(?<=Ziff[.]\\s)\\d+"),
              Abs = str_extract(., "(?<=Abs[.]\\s)\\d+"),
              Law = str_extract(., "\\w+$"))}

Result:

  Art Ziff  Abs  Law
1 110 <NA>    3 StGB
2  10 <NA>    1 StGB
3 122 <NA> <NA>  SVG
4   1    2 <NA>  UWG

Upvotes: 1

Related Questions