[R] How to format data for time-series analysis
jholtman at gmail.com
Tue Oct 23 00:40:29 CEST 2007
I would suggest that you use POSIXct as the format for the date/time.
This will give you resolution to almost a microsecond, so you should
be able to differentiate between multiple sequencial events. As for
reading it in, almost anything is reasonable; e.g., 2007/10/22
18:38:25.123456. You just have to make sure the sending system gives
you the time resolution that you need. "strptime" lets you convert
the character string to POSIXct.
As to your data format, usually each event is a record (line of text)
with the time and then event data.
On 10/22/07, B. Bogart <bbogart at sfu.ca> wrote:
> Hello all,
> I'm using R to visualize and explore the data produced by a software
> system. The software generates logs for many types of events. The
> software runs for days on end, and can possibly generate multiple events
> per second.
> What is the appropriate time format for year, month, day, hour, minute,
> second, millisecond? that R can properly interpret (using zoo? or ITS?)
> I'll have to change the software to generate the time tags in the
> appropriate format. Hopefully the format will not have any spaces so I
> can use "cut" to separate fields.
> Is it best practice to put each different type of event in a different
> field? What is the best data structure to hold this? I've only used
> "tseries" data without time-tags thus far, and am new to R and don't
> totally understand the type of data structures possible in R (if anyone
> can suggest a good tutorial for that please let me know).
> Thanks all,
> B. Bogart
> Simon Fraser Unversity
> R-help at r-project.org mailing list
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
+1 513 646 9390
What is the problem you are trying to solve?
More information about the R-help