[R] simple read in with zoo using POSIXlt

[Achim Zeileis]

On Sat, 14 Apr 2012, knavero wrote:

>
> Achim Zeileis-4 wrote
>>
>> You just need to declare that the index is in two columns (1 and 2) and
>> then provide a function that extracts a suitable object from it:
>>
>> read.zoo("test.txt", header = FALSE, index = 1:2,
>>    FUN = function(x, y) strptime(paste(x, y), "%d/%m/%Y %H:%M"))
>>
>> Use an additional as.POSIXct(...) around the strptime() call if you want
>> to use POSIXct instead of POSIXlt which is typically recommended.
>>
>> See vignette("zoo-read", package = "zoo") for more examples.
>> Z
>>
>
> Unfortunately, it's not working as I hoped for. Let me elaborate,

I don't know why you make this so complicated. Either use

read.zoo("test.txt", header = FALSE, sep = "\t",
   format = "%d/%m/%Y %H:%M", tz = "")

which yields a POSIXct time index. Alternatively, you can produce POSIXlt 
via strptime:

read.zoo("test.txt", header = FALSE, sep = "\t",
   FUN = function(x) strptime(x, "%d/%m/%Y %H:%M"))

The former is recommended for use in zoo.

> new code:
>
> http://pastebin.com/axpPB6M8
>
> So for this, I understand that the read in works very well with POSIXct, but
> I want to utilize the vectors contained with the POSIXlt class (wday, yday,
> mon, etc.). Here's how the POSIXct read.zoo looks like in the shell when
> copy pasted:
>
>> test = read.zoo("http://dl.dropbox.com/u/41922443/test.txt",
> +    header = FALSE, sep = "\t",
> +    FUN = function(idx) as.POSIXct(strptime(idx,
> +       format = fmt, tz = "PDT"), format = fmt, tz = "PDT"),
> +    colClasses = rep(c(NA, "numeric", "NULL"), c(1, 1, 0)),
> +    aggregate = tail1)
>> test
> 2010-01-07 00:15:00 2010-01-07 00:30:00 2010-01-07 00:45:00 2010-01-07
> 01:00:00
>           1333.620            1333.388            1335.343
> 1334.251
> 2010-01-07 01:15:00 2010-01-07 01:30:00 2010-01-07 01:45:00 2010-01-07
> 02:00:00
>           1331.589            1328.695            1329.151
> 1329.077
> 2010-01-07 02:15:00 2010-01-07 02:30:00
>           1327.649            1326.789
>
> This is good when you just eyeball it, HOWEVER, when the date/time is looked
> at by the machine, it doesn't see vectors that can be accessed, but the lame
> numerical/double that is the UTC time from 1960 or whatever in seconds.
> Proof of this is the following:
>
>> unclass(index(test))
> [1] 1262823300 1262824200 1262825100 1262826000 1262826900 1262827800
> [7] 1262828700 1262829600 1262830500 1262831400
> attr(,"tzone")
> [1] "PDT"
>
> Now with this code:
>
> http://pastebin.com/pr2X78sX
>
> This just pisses me off....let me elaborate...or well, eff it. I'll just
> copy paste and you'll get my point:
>
>> test = read.zoo("http://dl.dropbox.com/u/41922443/test.txt",
> +    header = FALSE, sep = "\t",
> +    FUN = function(idx) as.POSIXlt(strptime(idx,
> +       format = fmt, tz = "PDT"), format = fmt, tz = "PDT"),
> +    colClasses = rep(c(NA, "numeric", "NULL"), c(1, 1, 0)),
> +    aggregate = tail1)
>> test
>       0
> 1326.789
>
> Basically, what the hell is 0 and 1326.789 doing there?.....right?
>
>
> --
> View this message in context: http://r.789695.n4.nabble.com/simple-read-in-with-zoo-using-POSIXlt-tp4557138p4558038.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> 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.
>