I’m having a strange problem when I read a .csv file using read_csv. I’m afraid I don’t think I can produce a reproducible example because the issue may involve my current R/RStudio session and how they interact while reading a file, requiring the file and a similar set up. I don’t know for sure, but I’m leaning towards some issue with a text encoding mismatch, but I know little about this and so I seek the advanced advice of the stack-overflow hive mind.

In any case, here’s the behavior I have.

The tibble is named `fl’ and contains the FIPS codes for all the administrative districts in the U.S. I’m looking at a ‘Census Area’ in Alaska as the example, but I have other similar cases in the same tibble.

The following produces sensible output.

> fl %>% filter(str_detect(fl$NAME, 'Hoonah'))
# A tibble: 1 × 6
  FIPS  NAME                      STATEFPn COUNTYFPn STATEFP COUNTYFP
  <chr> <chr>                        <int>     <int> <chr>   <chr>   
 1 02105 Hoonah–Angoon Census Area        2       105 02      105     

But, if I do the following, by typing the whole NAME at the console prompt, I get nothing.

> fl %>% filter(NAME=='Hoonah-Angoon Census Area')
 # A tibble: 0 × 6
 # … with 6 variables: FIPS <chr>, NAME <chr>, STATEFPn <int>, COUNTYFPn > <int>, STATEFP <chr>,
 #   COUNTYFP <chr>
 # ℹ Use `colnames()` to see all variable names

However, if I copy-and-paste from the first output, it works and I get this.

> fl %>% filter(NAME=='Hoonah–Angoon Census Area')
# A tibble: 1 × 6
  FIPS  NAME                      STATEFPn COUNTYFPn STATEFP COUNTYFP
  <chr> <chr>                        <int>     <int> <chr>   <chr>   
1 02105 Hoonah–Angoon Census Area        2       105 02      105     

I have some suspicion that’s it’s about some sort of character encoding issue/mismatch between my RStudio session and what’s in the file, despite the fact that, to the best of my knowledge the file (as checked by guess_encoding()) and my session (as set in ‘file:save with encoding’ and then using ‘file:reopen with encoding’) both read ‘UTF-8’.

Any ideas about what is happening?