R 中不正确的字符串替换

Question

I need to replace awkward strings in R, specifically the times that are in a weird format. The data looks like this:

      Date |    Time | AmbientTemp
2000-01-01 | 11:00 a |          25
2000-01-01 | 11:30 a |        25.5 
2000-01-01 | 11:00 p |          20
2000-01-01 | 11:30 p |        19.5

The a and p mean AM and PM respectively (obviously).

lubridate and base R cannot convert these dates to a correct format. Thus, I turned to the cumbersome str_replace_all function (from package stringr ) to convert ALL my times in a large dataframe: >130000 records.

Example functions:

uploadDat$Time = str_replace_all(uploadDat$Time,"11:00 a","11:00")
uploadDat$Time = str_replace_all(uploadDat$Time,"11:00 p","23:00")

I changed the class of the times using as.character() before applying stringr 's functions.

The result is perfect except for the 11'o clock times (like above) that are converted as follow:

      Date |   Time | AmbientTemp
2000-01-01 | 101:00 |          25
2000-01-01 | 101:30 |        25.5 
2000-01-01 | 113:30 |          20
2000-01-01 | 113:30 |        19.5

Why are these specific times converted incorrectly?

Answer 1

We can paste "m" at the end of time, convert it into POSIXct

format(as.POSIXct(paste0(df$Time, "m"), format = "%I:%M %p"), "%T")
#[1] "11:00:00" "11:30:00" "23:00:00" "23:30:00"