Separating multi-valued attributes into individual attributes R - r

Question

I'm working with the stackoverflow developer survey data-set and attempting to predict compensation based on technologies worked with and collaborative tools worked with. These two attributes are multi-valued with semicolons separating the individual values.
For instance, under the CollabToolsWorkedWith attribute in one row, there is Confluence;Jira;Github;Slack;Microsoft;Teams;Google Suite. I want to give each of these values their own column with a value of either 0 or 1 if the row had that value.
The end result would have each row contain a column for every single value under CollabToolsWorkedWith and each column would contain 0's and 1's based on whether or not the row contained that value.

Answer 1

You may get a quicker answer next time if you provide some sample data that everyone can quickly access. I found the 2020 data online. Here is my answer:
# read the data frame
rm(list = ls())
df <- read.csv("survey_results_public.csv")
# figure out which column you are talking about
data.frame(colnames(df))
table(df$NEWCollabToolsWorkedWith)
# convert to lower case and character
df$NEWCollabToolsWorkedWith <- as.character(df$NEWCollabToolsWorkedWith)
df$NEWCollabToolsWorkedWith <- tolower(df$NEWCollabToolsWorkedWith)
# keep only the useful variables and separate based on ;
library(tidyverse)
library(splitstackshape)
namesdf <- df %>% select(NEWCollabToolsWorkedWith)
namesdf <- cSplit(namesdf,"NEWCollabToolsWorkedWith", sep = ";", direction = "wide", drop=TRUE,
type.convert = TRUE)
# stack stuff on top of each other to find unique list of tools/platforms
long_data_frame <-
namesdf %>%
pivot_longer(cols = starts_with("NEWCollabToolsWorkedWith"), # use columns starting with "year"
names_to ="unique", # name of new column
names_prefix = "_",
values_drop_na = TRUE) %>%
distinct(value)
# clean the variable names
library(janitor)
long_data_frame$value = as.character(long_data_frame$value)
long_data_frame$value = janitor::make_clean_names(long_data_frame$value)
# get final unique list
table(long_data_frame$value)
> table(long_data_frame$value)
confluence facebook_workplace github gitlab
1 1 1 1
google_suite_docs_meet_etc jira microsoft_azure microsoft_teams
1 1 1 1
slack stack_overflow_for_teams trello
1 1 1
# create new variables
df$confluence <- NA
df$jira <- NA
df$slack = NA
df$microsoft_azure =NA
df$trello = NA
df$github = NA
df$gitlab = NA
df$google_suite_docs_meet_etc = NA
df$microsoft_teams = NA
df$stack_overflow_for_teams = NA
df$facebook_workplace =NA
# make a dummy variable based on string match
df$confluence <- as.integer(grepl(pattern = "confluence", x = df$NEWCollabToolsWorkedWith))
df$jira <- as.integer(grepl(pattern = "jira", x = df$NEWCollabToolsWorkedWith))
df$slack <- as.integer(grepl(pattern = "slack", x = df$NEWCollabToolsWorkedWith))
df$microsoft_azure <- as.integer(grepl(pattern = "microsoft azure", x = df$NEWCollabToolsWorkedWith))
df$trello <- as.integer(grepl(pattern = "trello", x = df$NEWCollabToolsWorkedWith))
df$github <- as.integer(grepl(pattern = "github", x = df$NEWCollabToolsWorkedWith))
df$gitlab <- as.integer(grepl(pattern = "gitlab", x = df$NEWCollabToolsWorkedWith))
df$google_suite_docs_meet_etc <- as.integer(grepl(pattern = "google", x = df$NEWCollabToolsWorkedWith))
df$microsoft_teams <- as.integer(grepl(pattern = "microsoft teams", x = df$NEWCollabToolsWorkedWith))
df$stack_overflow_for_teams <- as.integer(grepl(pattern = "overflow", x = df$NEWCollabToolsWorkedWith))
df$facebook_workplace <- as.integer(grepl(pattern = "facebook", x = df$NEWCollabToolsWorkedWith))
# proof that it went through
table(df$facebook_workplace)
> table(df$facebook_workplace)
0 1
62881 1580