How to combine multiple character columns into a single column in an R data frame

Question

I am working with Census data and I need to combine four character columns into a single column.

Example:

LOGRECNO STATE COUNTY  TRACT BLOCK
    60           


        

Answer 1

Try this:

AL_Blocks$BLOCK_ID<- with(AL_Blocks, paste0(STATE, COUNTY, TRACT, BLOCK))

there was a typo in County... it should've been COUNTY. Also, you don't need the collapse parameter.

I hope that helps.

Answer 2

You can use do.call and paste0. Try:

AL_Blocks$BLOCK_ID <- do.call(paste0, AL_Block[c("STATE", "COUNTY", "TRACT", "BLOCK")])

Example output:

do.call(paste0, AL_Blocks[c("STATE", "COUNTY", "TRACT", "BLOCK")])
# [1] "010010211001053" "010010211001054" "010010211001055" "010010211001056"
# [5] "010010211001057" "010010211001058"
do.call(paste0, AL_Blocks[2:5])
# [1] "010010211001053" "010010211001054" "010010211001055" "010010211001056"
# [5] "010010211001057" "010010211001058"

---

You can also use unite from "tidyr", like this:

library(tidyr)
library(dplyr)
AL_Blocks %>% 
  unite(BLOCK_ID, STATE, COUNTY, TRACT, BLOCK, sep = "", remove = FALSE)
#   LOGRECNO        BLOCK_ID STATE COUNTY  TRACT BLOCK
# 1       60 010010211001053    01    001 021100  1053
# 2       61 010010211001054    01    001 021100  1054
# 3       62 010010211001055    01    001 021100  1055
# 4       63 010010211001056    01    001 021100  1056
# 5       64 010010211001057    01    001 021100  1057
# 6       65 010010211001058    01    001 021100  1058

where "AL_Blocks" is provided as:

AL_Blocks <- structure(list(LOGRECNO = c("60", "61", "62", "63", "64", "65"), 
    STATE = c("01", "01", "01", "01", "01", "01"), COUNTY = c("001", "001", 
    "001", "001", "001", "001"), TRACT = c("021100", "021100", "021100", 
    "021100", "021100", "021100"), BLOCK = c("1053", "1054", "1055", "1056",
    "1057", "1058")), .Names = c("LOGRECNO", "STATE", "COUNTY", "TRACT", 
    "BLOCK"), class = "data.frame", row.names = c(NA, -6L))

Answer 3

Or try this

DF$BLOCKID <-
  paste(DF$LOGRECNO, DF$STATE, DF$COUNTY, 
        DF$TRACT, DF$BLOCK, sep = "")

(Here is a method to set up the dataframe for people coming into this discussion later)

DF <- 
  data.frame(LOGRECNO = c(60, 61, 62, 63, 64, 65),
             STATE = c(1, 1, 1, 1, 1, 1),
             COUNTY = c(1, 1, 1, 1, 1, 1), 
             TRACT = c(21100, 21100, 21100, 21100, 21100, 21100), 
             BLOCK = c(1053, 1054, 1055, 1056, 1057, 1058))

Answer 4

You can both WRITE and READ Text files with any specified "string-separator", not necessarily a character separator. This is very useful in many cases when the data has practically all terminal symbols, and thus, no 1 symbol can be used as a separator. Here are examples of read and write functions:

WRITE OUT Special Separator Text:

writeSepText <- function(df, fileName, separator) {
    con <- file(fileName)
    data <- apply(df, 1, paste, collapse = separator)
    # data
    data <- writeLines(data, con)
    close(con)
    return
}

Test Writing out text file separated by a string "bra_break_ket"

writeSepText(df=as.data.frame(Titanic), fileName="/Users/user/break_sep.txt", separator="<break>")

READ In text files with special separator string

readSepText <- function(fileName, separator) {
    data <- readLines(con <- file(fileName))
    close(con)
    records <- sapply(data, strsplit, split=separator)
    dataFrame <- data.frame(t(sapply(records,c)))
    rownames(dataFrame) <- 1: nrow(dataFrame)
    return(as.data.frame(dataFrame,stringsAsFactors = FALSE))
}

Test Reading in text file separated by

df <- readSepText(fileName="/Users/user/break_sep.txt", separator="<break>"); df

Answer 5

You can use tidyverse package:

DF %>% unite(new_var, STATE, COUNTY, TRACT, BLOCK)

Answer 6

You can try this too

AL_Blocks <- transform(All_Blocks, BLOCKID = paste(STATE,COUNTY,
                       TRACT, BLOCK, sep = "")