[R-SIG-Finance] [Fwd: Re: stuck on data]

[Gabor Grothendieck]

Please provide ***reproducible*** examples: baseline.csv is missing
from your example -- a cut down version of it should be posted.
Also use dput(obj), not str(obj) when posting so that the output
is paste-able back into R.

The use of zoo() is wrong in your pass.   zoo() accepts a
numeric vector or matrix as arg 1 (or a factor) and a time index
as arg 2 but you are passing a data frame as arg 1 and a data frame
is not a numeric object.

Read the three zoo vignettes and note the error here is covered
in the FAQ:

vignette("zoo-faq")

Also note that there are numerous examples in the help files.  It
would be better to take a little time and go through this rather than
using trial and error without reference to the extensive documentation.

On Thu, Dec 25, 2008 at 12:40 PM, zubin <binabina at bellsouth.net> wrote:
>
>> Okay, i think i solved the dates issue with your help, but hung up
>> again...  My objective is to load this panel data set (cross sectional
>> time series data) into a zoo object so i can use the time series
>> functions(like arima, spectrum, etc..) and the quantmod library.
>> However, it seems you can't have panel data in these objects, only
>> multivariate time series (basically not data that is stacked or
>> pooled, so there are multiple dates by cross-section unit).   So what
>> i do is load the zoo object by store id#, which is okay - see below.
>> I can convert to zoo, then when i try to utilize problems occur.  Zoo
>> object converts all my integer data to character, i think??
>>
>>
>> >setwd("c:/csv")
>> >d1 <- read.csv("baseline.csv",stringsAsFactors=FALSE)
>> >dsmall <-
>> subset(d1[,c('total_Cases','date','Store_nbr')],Store_nbr=='177')
>> >dsmall$z <- as.Date(dsmall$date, "%d%b%Y")
>> >dsmallna <- na.omit(dsmall)
>> >dsmallna$date <- NULL
>> >str(dsmallna)
>>
>> 'data.frame':    414 obs. of  3 variables:
>>  $ total_Cases: int  246 6502 5718 6234 853 6387 9833 6138 2405 8066 ...
>>  $ Store_nbr  : int  177 177 177 177 177 177 177 177 177 177 ...
>>  $ z          :Class 'Date'  num [1:414] 13770 13771 13772 13773 13774 ...
>>  - attr(*, "na.action")=Class 'omit'  Named int [1:92] 415 416 417 418
>> 419 420 421 422 423 424 ...
>>   .. ..- attr(*, "names")= chr [1:92] "415" "416" "417" "418" ...
>>
>> this looks fine, the data.frame - total_Cases is variable of interest
>> and its an integer.
>>
>> >dzoo <- zoo(dsmallna, dsmallna$z)
>> >str(dzoo)
>>
>> 'zoo' series from 2007-09-14 to 2008-10-31
>>   Data: chr [1:414, 1:3] "  246" " 6502" " 5718" " 6234" "  853" ...
>>  - attr(*, "dimnames")=List of 2
>>   ..$ : chr [1:414] "1" "2" "3" "4" ...
>>   ..$ : chr [1:3] "total_Cases" "Store_nbr" "z"
>>   Index: Class 'Date'  num [1:414] 13770 13771 13772 13773 13774 ...
>>
>> /*okay, i don't know how to interpret the structure of the zoo object,
>> Data: chr?  seems like it shifted to character? */ /*How does one
>> interpret this object - $ : chr, $ : chr? */
>>
>> Now lets see if i can use the zoo object for some time series work...
>> Nope all these errors, i think related to the character conversion the
>> zoo object has done?
>>
>> > plot(dzoo)
>> i get a plot of total_Cases and store number and these errors?
>>
>> Error in plot.window(...) : need finite 'ylim' values
>> In addition: Warning messages:
>> 1: In xy.coords(x, y, xlabel, ylabel, log) : NAs introduced by coercion
>> 2: In min(x) : no non-missing arguments to min; returning Inf
>> 3: In max(x) : no non-missing arguments to max; returning -Inf
>>
>>
>> >chartSeries(dzoo)
>>
>> Error in checkSlotAssignment(object, name, value) : assignment of an
>> object of class "character" is not valid for slot "yrange" in an
>> object of class "chob"; is(value, "numeric") is not TRUE
>>
>>
>> >acf(dzoo)
>>
>> Error in acf(dzoo) : 'x' must be numeric
>>
>>
>> -zubin
>>
>>
>>
>
>        [[alternative HTML version deleted]]
>
> _______________________________________________
> R-SIG-Finance at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-finance
> -- Subscriber-posting only.
> -- If you want to post, subscribe first.
>