Regex: Checking rows with a specific pattern

library(readr)
library(tidyverse)
library(purrr)

adult <- read_csv("adultincome.csv")

## Parsed with column specification:
## cols(
##   age = col_double(),
##   workclass = col_character(),
##   fnlwgt = col_double(),
##   education = col_character(),
##   education.num = col_double(),
##   marital.status = col_character(),
##   occupation = col_character(),
##   relationship = col_character(),
##   race = col_character(),
##   sex = col_character(),
##   capital.gain = col_double(),
##   capital.loss = col_double(),
##   hours.per.week = col_double(),
##   native.country = col_character(),
##   income = col_character()
## )

We are looking for the total number of rows containing the patter ‘?’

missing_count_tbl <- purrr::map_df(adult, ~ stringr::str_detect(., pattern = "\\?")) %>%
  rowSums() %>%
  tbl_df() %>%
  filter(value > 0) %>%
  summarize(missing_count = n()) 

missing_count_tbl

## # A tibble: 1 x 1
##   missing_count
##           <int>
## 1          2399

In the data, there are 2399 rows that contain the pattern “?”.

To locate columns that have this pattern

count.NA.percolumn <- plyr::ldply(adult, function(c) sum(c == "?"))
count.NA.percolumn

##               .id   V1
## 1             age    0
## 2       workclass 1836
## 3          fnlwgt    0
## 4       education    0
## 5   education.num    0
## 6  marital.status    0
## 7      occupation 1843
## 8    relationship    0
## 9            race    0
## 10            sex    0
## 11   capital.gain    0
## 12   capital.loss    0
## 13 hours.per.week    0
## 14 native.country  583
## 15         income    0

There are 3 columns that contain “?” as NA: workclass, occupation, and native.country.