Using gsub to create uniform time formats

Question

I am trying to reformat times that are currently in a character format. I'm trying to make them uniform. Right now they look like this:

 [1] "1:00PM"      "1:10 PM"     "1:10PM"      "1:20 PM"     "1:30 PM"     "1:30PM"     
 [7] "1:40 PM"     "10:00AM"     "10:10 AM"    "10:10AM"     "10:30 AM"    "10:30AM"    
 [13] "10:45 AM"    "10:45AM"     "10:50 AM"    "10:50AM"     "10AM"        "11:00AM"    
 [19] "11:10 AM"    "11:10AM"     "11:40 AM"    "11:40AM"     "11AM"        "12:00PM"    
 [25] "12:05 PM"    "12:10 PM"    "12:10PM"     "12:25PM"     "12:30 PM"    "12:30PM"    
 [31] "12:45 PM"    "12:45:30 PM" "12:45PM"     "12:50 PM"    "12PM"        "1PM"        
 [37] "2:00PM"      "2:10 PM"     "2:10PM"      "2:20PM"      "2:30 PM"     "2:30PM"     
 [43] "2:35 PM"     "2:45 PM"     "2:45PM"      "2:55 PM"     "2PM"         "3:00PM"     
 [49] "3:05 PM"     "3:10 PM"     "3:10PM"      "3:20 PM"     "3:20PM"      "3:25 PM"    
 [55] "3:25PM"      "3:30 PM"     "3:35 PM"     "3:35PM"      "3:45 PM"     "3:45PM"     
 [61] "3PM"         "4:00PM"      "4:10 PM"     "4:10PM"      "4:30 PM"     "4:30PM"     
 [67] "4:35 PM"     "4:35PM"      "4PM"         "5:00PM"      "5:10 PM"     "5:10PM"     
 [73] "5:20 PM"     "5:30 PM"     "5:30PM"      "5:35 PM"     "5:35PM"      "5:40 PM"    
 [79] "5:40PM"      "5:45 PM"     "5:50 PM"     "5:50PM"      "6:00PM"      "6:05PM"     
 [85] "6:10 PM"     "6:10PM"      "6:15PM"      "6:30 PM"     "6:30PM"      "6PM"        
 [91] "7:00PM"      "7:10 AM"     "7:10 PM"     "7:10AM"      "7:10PM"      "7:30PM"     
 [97] "7:35 PM"     "7:35PM"      "7:45 PM"     "7:45PM"      "7AM"         "7PM"        
[103] "8:00AM"      "8:10 AM"     "8:10AM"      "8:25 PM"     "8:25PM"      "8:50 PM"    
[109] "8AM"         "9:00AM"      "9:10 AM"     "9:10AM"      "9:15 AM"     "9:15AM"     
[115] "9:20 AM"     "9:30 AM"     "9:30AM"      "9:35AM"      "9:40 AM"     "9:40AM"     
[121] "9:45 AM"     "9:45AM"      "9AM"

I want all times to be in this format:

1:00PM instead of 1PM
12:45PM instead of 12:45:30 PM

So basically HH:MM appended with AM or PM

Eventually I would like to convert the times from character format into POSIXct format. But that is only possible with a uniform character format. To be more specific: how would you use gsub to change "3PM" into "3:00PM" and likewise change "12:45:30 PM" into "12:45PM", as an example?

I have a hard time wrapping my head around some of the Regular Expression Syntax in gsub - especially how to reference a specific position, say position 4 in a character string.

http://stackoverflow.com/questions/1803627/understanding-dates-times-posixc-and-posixct-in-r — jmargolisvt, Jun 07 '15 at 03:53

akrun · Accepted Answer · 2015-06-07T07:03:24.640

We create a index ('indx') for elements that don't have : i.e. (10AM, 11AM etc.), using sub we change the format (10:00AM, 11:00AM etc.). We match the first two digits, followed by :, two digits (\\d{2}), capture that as a group with parentheses, match characters that are not AM/PM ([^AMP]+), match characters that are AM/PM and select as the 2nd capture group, use the first and second captured groups (\\1\\2) as replacement. Now we can either use strsplit/sprintf to pad 0s at the beginning for elements that don't have two digits.

indx <- !grepl(':', str1)
str1[indx] <- sub('(\\d+)(.*)', '\\1:00\\2', str1[indx])
str1 <- sub('(^\\d+:\\d{2})[^AMP]+([AMP])', '\\1\\2', str1)
sapply(strsplit(str1, ':'), function(x) paste(sprintf('%02d',
          as.numeric(x[1])), x[2], sep=":"))
#[1] "01:00PM" "01:10PM" "01:10PM" "01:20PM" "01:30PM" "01:30PM" "01:40PM"
#[8] "10:00AM" "10:10AM" "10:10AM" "10:30AM" "10:30AM" "10:45AM" "10:45AM"
#[15] "10:50AM" "10:50AM" "10:00AM" "11:00AM" "11:10AM" "11:10AM" "11:40AM"
#[22] "11:40AM" "11:00AM" "12:00PM" "12:05PM" "12:10PM" "12:10PM" "12:25PM"
#[29] "12:30PM" "12:30PM" "12:45PM" "12:45PM" "12:45PM" "12:50PM" "12:00PM"
#[36] "01:00PM" "02:00PM" "02:10PM" "02:10PM" "02:20PM" "02:30PM" "02:30PM"
#[43] "02:35PM" "02:45PM" "02:45PM" "02:55PM" "02:00PM" "03:00PM" "03:05PM"
#[50] "03:10PM" "03:10PM" "03:20PM" "03:20PM" "03:25PM" "03:25PM" "03:30PM"
#[57] "03:35PM" "03:35PM" "03:45PM" "03:45PM" "03:00PM" "04:00PM" "04:10PM"
#[64] "04:10PM" "04:30PM" "04:30PM" "04:35PM" "04:35PM" "04:00PM" "05:00PM"
#[71] "05:10PM" "05:10PM" "05:20PM" "05:30PM" "05:30PM" "05:35PM" "05:35PM"
#[78] "05:40PM" "05:40PM" "05:45PM" "05:50PM" "05:50PM" "06:00PM" "06:05PM"
#[85] "06:10PM" "06:10PM" "06:15PM" "06:30PM" "06:30PM" "06:00PM" "07:00PM"
#[92] "07:10AM" "07:10PM" "07:10AM" "07:10PM" "07:30PM" "07:35PM" "07:35PM"
#[99] "07:45PM" "07:45PM" "07:00AM" "07:00PM" "08:00AM" "08:10AM" "08:10AM"
#[106] "08:25PM" "08:25PM" "08:50PM" "08:00AM" "09:00AM" "09:10AM" "09:10AM"
#[113] "09:15AM" "09:15AM" "09:20AM" "09:30AM" "09:30AM" "09:35AM" "09:40AM"
#[120] "09:40AM" "09:45AM" "09:45AM" "09:00AM"

Or instead of strsplit and sapply, we can use format

 sub('^ ', '0', format(str1, justify='right'))

Or

 library(stringr)
 str_pad(str1, pad='0', width=7)

Or we can use lubridate package which has options for multiple format strings

 library(lubridate)
 paste0(format(parse_date_time(str1, orders=guess_formats(gsub('[APM]', 
   '', str1), c('hm', 'hms', 'h'))), '%H:%M'), sub('[^AMP]+', '', str1))
 #[1] "01:00PM" "01:10PM" "01:10PM" "01:20PM" "01:30PM" "01:30PM" "01:40PM"
 #[8] "10:00AM" "10:10AM" "10:10AM" "10:30AM" "10:30AM" "10:45AM" "10:45AM"
 #[15] "10:50AM" "10:50AM" "10:00AM" "11:00AM" "11:10AM" "11:10AM" "11:40AM"
 #[22] "11:40AM" "11:00AM" "12:00PM" "12:05PM" "12:10PM" "12:10PM" "12:25PM"
 #[29] "12:30PM" "12:30PM" "12:45PM" "12:45PM" "12:45PM" "12:50PM" "12:00PM"
 #[36] "01:00PM" "02:00PM" "02:10PM" "02:10PM" "02:20PM" "02:30PM" "02:30PM"
 #[43] "02:35PM" "02:45PM" "02:45PM" "02:55PM" "02:00PM" "03:00PM" "03:05PM"
 #[50] "03:10PM" "03:10PM" "03:20PM" "03:20PM" "03:25PM" "03:25PM" "03:30PM"
 #[57] "03:35PM" "03:35PM" "03:45PM" "03:45PM" "03:00PM" "04:00PM" "04:10PM"
 #[64] "04:10PM" "04:30PM" "04:30PM" "04:35PM" "04:35PM" "04:00PM" "05:00PM"
 #[71] "05:10PM" "05:10PM" "05:20PM" "05:30PM" "05:30PM" "05:35PM" "05:35PM"
 #[78] "05:40PM" "05:40PM" "05:45PM" "05:50PM" "05:50PM" "06:00PM" "06:05PM"
 #[85] "06:10PM" "06:10PM" "06:15PM" "06:30PM" "06:30PM" "06:00PM" "07:00PM"
 #[92] "07:10AM" "07:10PM" "07:10AM" "07:10PM" "07:30PM" "07:35PM" "07:35PM"
 #[99] "07:45PM" "07:45PM" "07:00AM" "07:00PM" "08:00AM" "08:10AM" "08:10AM"
 #[106] "08:25PM" "08:25PM" "08:50PM" "08:00AM" "09:00AM" "09:10AM" "09:10AM"
 #[113] "09:15AM" "09:15AM" "09:20AM" "09:30AM" "09:30AM" "09:35AM" "09:40AM"
 #[120] "09:40AM" "09:45AM" "09:45AM" "09:00AM"

data

str1 <- c("1:00PM", "1:10 PM", "1:10PM", "1:20 PM", "1:30 PM", "1:30PM", 
"1:40 PM", "10:00AM", "10:10 AM", "10:10AM", "10:30 AM", "10:30AM", 
"10:45 AM", "10:45AM", "10:50 AM", "10:50AM", "10AM", "11:00AM", 
"11:10 AM", "11:10AM", "11:40 AM", "11:40AM", "11AM", "12:00PM", 
"12:05 PM", "12:10 PM", "12:10PM", "12:25PM", "12:30 PM", "12:30PM", 
"12:45 PM", "12:45:30 PM", "12:45PM", "12:50 PM", "12PM", "1PM", 
"2:00PM", "2:10 PM", "2:10PM", "2:20PM", "2:30 PM", "2:30PM", 
"2:35 PM", "2:45 PM", "2:45PM", "2:55 PM", "2PM", "3:00PM", "3:05 PM", 
"3:10 PM", "3:10PM", "3:20 PM", "3:20PM", "3:25 PM", "3:25PM", 
"3:30 PM", "3:35 PM", "3:35PM", "3:45 PM", "3:45PM", "3PM", "4:00PM", 
"4:10 PM", "4:10PM", "4:30 PM", "4:30PM", "4:35 PM", "4:35PM", 
"4PM", "5:00PM", "5:10 PM", "5:10PM", "5:20 PM", "5:30 PM", "5:30PM", 
"5:35 PM", "5:35PM", "5:40 PM", "5:40PM", "5:45 PM", "5:50 PM", 
"5:50PM", "6:00PM", "6:05PM", "6:10 PM", "6:10PM", "6:15PM", 
"6:30 PM", "6:30PM", "6PM", "7:00PM", "7:10 AM", "7:10 PM", "7:10AM", 
"7:10PM", "7:30PM", "7:35 PM", "7:35PM", "7:45 PM", "7:45PM", 
"7AM", "7PM", "8:00AM", "8:10 AM", "8:10AM", "8:25 PM", "8:25PM", 
"8:50 PM", "8AM", "9:00AM", "9:10 AM", "9:10AM", "9:15 AM", "9:15AM", 
"9:20 AM", "9:30 AM", "9:30AM", "9:35AM", "9:40 AM", "9:40AM", 
"9:45 AM", "9:45AM", "9AM")

The first solution, using **strsplit** and **sapply**, worked perfectly. Thanks a lot for the elegant solution. — WykoW, Jun 07 '15 at 07:37

Using gsub to create uniform time formats

1 Answers1

data