R help - Sep 2010 - Splitting a time + duration into a series of periods

Hi all,

Here's what I have. I have a list of log-in times for users and how
long their sessions were.

user, login_time, session_min
alice, 2010/01/01 10:00, 145
bob, 2010/01/01 11:00, 30

What I want to do is create a bar chart showing, in 1/2 hour segments,
the number of users logged in at the same time.  For example:

10:00 1
10:30 1
11:00 2
11:30 1

The only way I can do this now is to send the data through a Perl
script to generate raw data like:

alice, 2010/01/01, 10:00
alice, 2010/01/01, 10:30
alice, 2010/01/01, 11:00
...
bob, 2010/01/01, 11:00

I've looked through the man pages for a couple hours now, and I can't
find a better of way of doing this directly in R.  Any help or
pointers?  Thanks in advance.

Travers -
    R's date/time abilities are pretty powerful -- you shouldn't have to
resort to outside programs.  Here's how I'd approach the problem:
> dat = read.csv(textConnection('user, login_time, session_min
+ alice, 2010/01/01 10:00, 145
+ bob, 2010/01/01 11:00, 30
+ '),stringsAsFactors=FALSE)
> dat$login_time = as.POSIXct(dat$login_time)
> dat$login_end = dat$login_time + 60 * dat$session_min
> tms = seq(as.POSIXct('2010/01/01 10:00'),
+           as.POSIXct('2010/01/01
12:30'),by=30*60)
> data.frame(time=tms,count=sapply(tms,
+             function(tm)sum(dat$login_time <= tm & dat$login_end >
tm)))
                  time count
1 2010-01-01 10:00:00     1
2 2010-01-01 10:30:00     1
3 2010-01-01 11:00:00     2
4 2010-01-01 11:30:00     1
5 2010-01-01 12:00:00     1
6 2010-01-01 12:30:00     0

Hope this helps.

 					- Phil Spector
 					 Statistical Computing Facility
 					 Department of Statistics
 					 UC Berkeley
 					 spector at stat.berkeley.edu



On Fri, 24 Sep 2010, Travers Naran wrote:
> Hi all,
>
> Here's what I have. I have a list of log-in times for users and how
> long their sessions were.
>
> user, login_time, session_min
> alice, 2010/01/01 10:00, 145
> bob, 2010/01/01 11:00, 30
>
> What I want to do is create a bar chart showing, in 1/2 hour segments,
> the number of users logged in at the same time.  For example:
>
> 10:00 1
> 10:30 1
> 11:00 2
> 11:30 1
>
> The only way I can do this now is to send the data through a Perl
> script to generate raw data like:
>
> alice, 2010/01/01, 10:00
> alice, 2010/01/01, 10:30
> alice, 2010/01/01, 11:00
> ...
> bob, 2010/01/01, 11:00
>
> I've looked through the man pages for a couple hours now, and I
can't
> find a better of way of doing this directly in R.  Any help or
> pointers?  Thanks in advance.
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

Try this:

findInterval(do.call(seq, c(as.list(as.POSIXct(DF$login_time)), by = '30
mins')), as.POSIXct(DF$login_time))



On Fri, Sep 24, 2010 at 4:32 PM, Travers Naran <tnaran@gmail.com> wrote:
> Hi all,
>
> Here's what I have. I have a list of log-in times for users and how
> long their sessions were.
>
> user, login_time, session_min
> alice, 2010/01/01 10:00, 145
> bob, 2010/01/01 11:00, 30
>
> What I want to do is create a bar chart showing, in 1/2 hour segments,
> the number of users logged in at the same time.  For example:
>
> 10:00 1
> 10:30 1
> 11:00 2
> 11:30 1
>
> The only way I can do this now is to send the data through a Perl
> script to generate raw data like:
>
> alice, 2010/01/01, 10:00
> alice, 2010/01/01, 10:30
> alice, 2010/01/01, 11:00
> ...
> bob, 2010/01/01, 11:00
>
> I've looked through the man pages for a couple hours now, and I
can't
> find a better of way of doing this directly in R.  Any help or
> pointers?  Thanks in advance.
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>


-- 
Henrique Dallazuanna
Curitiba-Paraná-Brasil
25° 25' 40" S 49° 16' 22" O

	[[alternative HTML version deleted]]