用于过滤掉 R 数据框列的重复数字的正则表达式

Question

I have a data frame with many columns and rows and I need to filter based on the value of two columns (Lat and Lon). I need a regular expression which

1. Removes any row for which either the Lat or Lon column does not have at least three decimal places. So the first row (human) would be filtered, because even though Lon has three decimal places, Lat does not.
2. Removes any row for which the decimal places are redundant. What I mean by redundant is there are three repeats of the same number continuing to the end. But if the redundancy starts after the third decimal, it doesn't matter. And if the redundancy is eventually followed by a different number, it doesn't matter.

Type <-c("human","camera","ebird","museum", "specimen", "gbif")
Lat <- c(34.67, 34.66,34.6666666, 34.666582, 34.56666, 34.586666)
Lon <- c(9.888,9.88,9.8761,9.888064, 9.78888,9.318888)
x = data.frame(cbind(Type,Lat,Lon))

Here's how each row would fare under the regex:

1. fails because Lat only has two decimal places, even though Lon passes.
2. fails because both rows only have two decimal places
3. fails because Lat repeats the same value, starting at the first decimal place, and the repetition continues to the end of the number.
4. Passes the regex
5. Fails because the repetitive number values starts at the second decimal places and continues for at least 3 repetitions all the way to the end
6. Passes the regex

So the resulting data frame from this regex filter would be:

Type <-c("museum","gbif")
Lat <- c(34.666582, 34.586666)
Lon <- c(9.888064, 9.318888)
x = data.frame(cbind(Type,Lat,Lon))

Answer 1

The function below will output the desired dataframe that you want. It accomplishes all of the requirements you stated above.

check.expressions <- function(data){
  data$pass <- FALSE
  for(i in 1:nrow(data)){
    if(nchar(str_extract(x$Lon[i], "(?<=\\.).*")) < 3 | nchar(str_extract(x$Lat[i], "(?<=\\.).*")) < 3){
     next 
    } else {
      unlist(str_split(str_extract(x$Lon[i], "(?<=\\.).*" ), "")) -> lon
      unlist(str_split(str_extract(x$Lat[i], "(?<=\\.).*" ), "")) -> lat
      if(lon[1] == lon[2] && lon[2] == lon[3]){
        if(length(lon) > 3){
          if(lon[3] != lon[length(lon)]){
            data$pass[i] <- TRUE
            next
          } else {
            next
          }
        }
        next
      }
      if(lat[1] == lat[2] && lat[2] == lat[3]){
        if(length(lat) > 3){
          if(lat[3] != lat[length(lat)]){
            data$pass[i] <- TRUE
            next
          } else {
            next
          }
        }
        next
      }
      if(length(lon) > 4){
        if(lon[2] == lon[3] && lon[3] == lon[4]){
          if(lon[4] != lon[length(lon)]){
            data$pass[i] <- TRUE
            next
          } else {
            next
          }
        }
      }
      if(length(lat) > 4){
        if(lat[2] == lat[3] && lat[3] == lat[4]){
          if(lat[4] != lat[length(lat)]){
            data$pass[i] <- TRUE
            next
          }
        }
      }
      data$pass[i] <- TRUE
    }
  }
  data[data$pass == TRUE, ] -> data
  return(data)
}

The function call being just:

check.expressions(x) -> x.out

which would produce:

> check.expressions(x) -> x.out
> x.out
    Type       Lat      Lon pass
4 museum 34.666582 9.888064 TRUE
6   gbif 34.586666 9.318888 TRUE