Date time am/pm in R - r

Question

I'm having an issue with the datetime field of a timeseries:
> CO1temp[163:169,]
Date OPEN HIGH LOW CLOSE
163 7/11/2011 11:45:00 PM 116.30 116.30 116.09 116.18
164 7/11/2011 11:50:00 PM 116.16 116.78 116.13 116.70
165 7/11/2011 11:55:00 PM 116.69 116.83 116.51 116.65
166 7/12/2011 116.65 116.79 116.44 116.50
167 7/12/2011 12:05:00 AM 116.50 116.60 116.39 116.47
168 7/12/2011 12:10:00 AM 116.49 116.55 116.38 116.52
169 7/12/2011 12:15:00 AM 116.52 116.67 116.39 116.44
As you can see the midnight time (line 166) is not showing properly.
Which creates a NA when I create my xts object:
CO1 <- as.xts(CO1temp[, 2:5], order.by = as.POSIXct(CO1temp[,1],format='%m/%d/%Y %r'),frequency="5 minutes")
> CO1[163:169,]
OPEN HIGH LOW CLOSE
2011-07-11 23:45:00 116.30 116.30 116.09 116.18
2011-07-11 23:50:00 116.16 116.78 116.13 116.70
2011-07-11 23:55:00 116.69 116.83 116.51 116.65
<NA> 116.65 116.79 116.44 116.50
2011-07-12 00:05:00 116.50 116.60 116.39 116.47
2011-07-12 00:10:00 116.49 116.55 116.38 116.52
2011-07-12 00:15:00 116.52 116.67 116.39 116.44
This later leads to more problem when I want to analyze this timeseries.
?strptime is quite specific about it:
The default for the format methods is "%Y-%m-%d %H:%M:%S" if any component has a time component which is not midnight, and "%Y-%m-%d" otherwise.
However my datetime is not in the standard format.
I would greatly appreciate any help.

Answer 1

This a kind of a hack but it works.
You just have to append "12:00:00 AM" to your vector of date: those which are lacking the hour information will be read correctly, and in the dates that already have the hour information it will just be ignored and only the one that was already there will be read.
CO1 <- as.xts(CO1temp[, 2:5],
order.by = as.POSIXct(paste(CO1temp$Date,"12:00:00 AM", sep=" "),
format='%m/%d/%Y %r'),
frequency="5 minutes")
CO1
OPEN HIGH LOW CLOSE
2011-07-11 23:45:00 116.30 116.30 116.09 116.18
2011-07-11 23:50:00 116.16 116.78 116.13 116.70
2011-07-11 23:55:00 116.69 116.83 116.51 116.65
2011-07-12 00:00:00 116.65 116.79 116.44 116.50
2011-07-12 00:05:00 116.50 116.60 116.39 116.47
2011-07-12 00:10:00 116.49 116.55 116.38 116.52
2011-07-12 00:15:00 116.52 116.67 116.39 116.44
That being said, if you ended up with your dataframe as it is after using strptime then your date column is already in POSIXct format and therefore the following should work directly:
as.xts(CO1temp[, 2:5], order.by = CO1temp$Date, frequency = "5 minutes")