Merging string variables

General
, ,
1

Hi,
I am trying to merge all string variables including comments longer than 3 characters so URN, Q6, Score and Year would be excluded in this df:

source <- data.frame(
  stringsAsFactors = FALSE,
                                    URN = c("H732585",
                                            "H933818","H902416",
                                            "H793061","H592160","H972119",
                                            "H945230","H955322",
                                            "H814977","H972992"),
                                   QN3a = c("dsafs",NA,"aaa",
                                            "djdjd fdfj","fff",
                                            "f  d ffsa j dsf aa sd","dffg ddjfcj mj",
                                            "dkvvf ffk vjf fj","ttt","fddd"),
                                   QN3b = c(NA,NA,"nil",NA,NA,
                                            "no comments",NA,NA,"all good",NA),
                                   QN5a = c("xxxxx","Nothing at all",
                                            "I did not have any","Non",NA,
                                            "Nothing","N/A","None",NA,
                                            "Nothing really"),
                                   QN5b = c("All good",NA,NA,NA,
                                            "nothing",NA,NA,"na","daa ffss fssfsff sfasfa",NA),
                                   Q6 = c("Yes","No",NA,NA,
                                  "Yes",NA,NA,"No",NA,NA),
                                  Score = c(100,90,35,20,50,90,
                                            100,100,90,80),
                                   Year = c(2021,2021,2020,2020,
                                            2021,2021,2021,2021,2020,2020)
                     )
library(dplyr)
library(stringr)
library(tidyr)

result <-  source %>% 
  mutate_at(vars(matches("QN5$|QN3$")), ~str_remove_all(.x, "^.{1,5}$")) %>% # Remove sentences with less than 5 characters
  mutate_at(vars(matches("QN5$|QN3$")), ~str_remove_all(.x, "^(All//sgood|No\\scomments|N.?A|Nothing|None|Nil)$")) %>% # Remove sentences with no comments
  mutate(all_comments = paste(QN3a,QN3b,QN5a,QN5b, sep="/"),
         all_comments = str_remove_all(all_comments, "NA"), # Removes NAs
         all_comments = str_remove_all(all_comments, "[:cntrl:]"), # Removes control characters like /n/r
         all_comments = str_replace_all(all_comments, "\\s\\s+", " "),  #Removes duplicated /
         all_comments = str_replace_all(all_comments, "//+", "/"), # Removes extra spaces
         all_comments = str_remove (all_comments, "/$"), # Removes / in the end
         all_comments = str_remove (all_comments, "^/"))  %>% # Removes / in the beginning
  mutate(All_len = nchar(all_comments),
         All_wcount = str_count(all_comments,'\\w+'))

Unfortunately:

  1. My str_remove_all rules do not work
  2. I don't know how to specify a code to take into account str with comments longer than 10 characters, rather then specifying them clearly in my code (QN3a,QN3b,QN5a,QN5b)

Can you help?

2

this wont evaluate to any source variable . none of them end with QN5, nor QN3. perhaps choose that starts with symbol for regex ^, or simply drop the $ end of line regex.

3

Thank you but I was looking for something like this:

paste where(~ is.character(.) & any(. > 10 , na.rm = TRUE))

so taking into account character variables with text longer than 10 characters.
I used a shortcut to take into account all variables start5ing from QN3 or QN5.

Also, I don't know why my str_remove_all commands do not work...

4

But I'm saying you didn't.

5

so is it not the above?

6

evidence

library(tidyverse) 

(qn5_is_involved <- data.frame(AQN5=1:3,
                              QN5A=1:3))
qn5_is_involved%>% 
  mutate(across(matches("QN5$"),~"X"))


qn5_is_involved %>% 
  mutate(across(matches("^QN5"),~"X"))
7

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.