mutate with gsub to clean out commas from a number - table scraped with rvest - r

Question

I'm practising scraping and data cleaning, and have a table which I've scraped from wikipedia. I'm trying to mutate the table to create a column which cleans out the commas from an existing column to return the number. All I'm getting is a column of NAs.
This is my output:
> library(dplyr)
> library(rvest)
>
> pg <- read_html("https://en.wikipedia.org/wiki/Rugby_World_Cup")
> rugby <- pg %>% html_table(., fill = T)
>
> rugby_table <- rugby[[3]]
>
> rugby_table
# A tibble: 9 x 8
Year `Host(s)` `Total attend­ance` Matches `Avg attend­ance` `% change in avg att.` `Stadium capacity` `Attend­ance as % o~
<int> <chr> <chr> <chr> <chr> <chr> <chr> <chr>
1 1987 Australia New Zealand 604,500 32 20,156 — 1,006,350 60%
2 1991 England France Ireland Scotland Wales 1,007,760 32 31,493 +56% 1,212,800 79%
3 1995 South Africa 1,100,000 32 34,375 +9% 1,423,850 77%
4 1999 Wales 1,750,000 41 42,683 +24% 2,104,500 83%
5 2003 Australia 1,837,547 48 38,282 –10% 2,208,529 83%
6 2007 France 2,263,223 48 47,150 +23% 2,470,660 92%
7 2011 New Zealand 1,477,294 48 30,777 –35% 1,732,000 85%
8 2015 England 2,477,805 48 51,621 +68% 2,600,741 95%
9 2019 Japan 1,698,528 45† 37,745 –27% 1,811,866 90%
>
> rugby_table2 <- rugby %>%
+ .[[3]] %>%
+ tbl_df %>%
+ mutate(Attendance=as.numeric(gsub("[^0-9.-]+","",'Total attendance')))
>
> rugby_table2
# A tibble: 9 x 9
Year `Host(s)` `Total attend­ance` Matches `Avg attend­ance` `% change in avg~ `Stadium capaci~ `Attend­ance as~ Attendance
<int> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <dbl>
1 1987 Australia New Zealand 604,500 32 20,156 — 1,006,350 60% NA
2 1991 England France Ireland Scotland Wales 1,007,760 32 31,493 +56% 1,212,800 79% NA
3 1995 South Africa 1,100,000 32 34,375 +9% 1,423,850 77% NA
4 1999 Wales 1,750,000 41 42,683 +24% 2,104,500 83% NA
5 2003 Australia 1,837,547 48 38,282 –10% 2,208,529 83% NA
6 2007 France 2,263,223 48 47,150 +23% 2,470,660 92% NA
7 2011 New Zealand 1,477,294 48 30,777 –35% 1,732,000 85% NA
8 2015 England 2,477,805 48 51,621 +68% 2,600,741 95% NA
9 2019 Japan 1,698,528 45† 37,745 –27% 1,811,866 90% NA
Any ideas?

Answer 1

The difficulty here is that gsub is interpreting 'Total attendance' as a character string, not a column name. My natural reaction was to use backticks instead of single quotes, but then I get a message that this object could not be found. I'm not sure what the problem is here, but you can resolve it using across
rugby_table2 <- rugby_table %>%
mutate(Attendance = across(contains("Total"),
function(x) as.numeric(gsub(",", "", x))),
Attendance = Attendance[[1]])
rugby_table2$Attendance
#> [1] 604500 1007760 1100000 1750000 1837547 2263223 1477294 2477805 1698528
EDIT
Ronak Shah has identified the problem, which is that there is an invisible character in the name brought across from the web page, which means the column isn't recognised. So an alternative solution would be:
names(rugby_table)[3] <- "Total attendance"
rugby_table2 <- rugby_table %>%
mutate(Attendance = as.numeric(gsub(",", "", `Total attendance`)))
rugby_table2$Attendance
#> [1] 604500 1007760 1100000 1750000 1837547 2263223 1477294 2477805

Answer 2

The gsub function is to do replacements to all matches of the provided pattern. If you are going to remove all the commas with gsub, the correct syntax would be
rugby_table2 <- rugby %>%
.[[3]] %>%
tbl_df %>%
mutate(Attendance = as.numeric(gsub(",", "", 'Total attendance')))
Edit:
rugby_table <- structure(list(Year = c(1987L, 1991L, 1995L, 1999L, 2003L, 2007L,
2011L, 2015L, 2019L), `Host(s)` = c("AustraliaNewZealand", "EnglandFranceIrelandScotlandWales",
"SouthAfrica", "Wales", "Australia", "France", "NewZealand",
"England", "Japan"), `Total attendance` = c("604,500", "1,007,760",
"1,100,000", "1,750,000", "1,837,547", "2,263,223", "1,477,294",
"2,477,805", "1,698,528"), Matches = c("32", "32", "32", "41",
"48", "48", "48", "48", "45+"), `Avg attendance` = c("20,156",
"31,493", "34,375", "42,683", "38,282", "47,150", "30,777", "51,621",
"37,745"), `% change in avg att` = c("—", "56%", "9%", "24%",
"–10%", "23%", "–35%", "68%", "–27%"), `Stadium capacity` = c("1,006,350",
"1,212,800", "1,423,850", "2,104,500", "2,208,529", "2,470,660",
"1,732,000", "2,600,741", "1,811,866"), `Attendance as % o~` = c("60%",
"79%", "77%", "83%", "83%", "92%", "85%", "95%", "90%")), row.names = c(NA,
-9L), class = c("tbl_df", "tbl", "data.frame"))
library(dplyr)
rugby_table %>%
mutate(Attendance = as.numeric(gsub(",", "", `Total attendance`))) %>%
select(Attendance)
#> # A tibble: 9 x 1
#> Attendance
#> <dbl>
#> 1 604500
#> 2 1007760
#> 3 1100000
#> 4 1750000
#> 5 1837547
#> 6 2263223
#> 7 1477294
#> 8 2477805
#> 9 1698528