[R] R: Help xts object Subset Date by Day of the Week

[R. Michael Weylandt]

On Sun, Aug 5, 2012 at 4:49 PM, Douglas Karabasz
<douglas at sigmamonster.com> wrote:
> I have a xts object made of daily closing prices I have acquired using
> quantmod.
>
>
>
> Here is my code:
>
> library(xts)
>
> library(quantmod)
>
> library(lubridate)
>
>
>
> # Gets SPY data
>
> getSymbols("SPY")
>
> # Subset Prices to just closing price
>
> SP500 <- Cl(SPY)
>
> # Show day of the week for each date using 2-6 for monday-friday
>
> SP500wd <- wday(SP500)
>
> # Add Price and days of week together
>
> SP500wd <- cbind(SP500, SP500wd)
>
> # subset Monday into one xts object
>
> SPmon <- subset(SP500wd, SP500wd$..2=="2")
>
>
>
>
>
> I then used the package lubridate to show the days of the week.   Due to the
> requirement of an xts objects to be numeric you will see each day is
> represented as a number so that Monday is =2, Tuesday=3, Wednesday=4,
> Thursday=5, Friday=6, Saturday=7.   Since this is a financial index you will
> only see the numbers 2-6 or Monday-Friday.
>
> I want to subset the data by using the day column.  I would like some help
> to figure out the best way to accomplish a few objectives.
>
> 1.       Subset the data so that I only show Monday in sequence.  However, I
> do want to make sure that it shows the date, price and the ..2 colum(which
> is the day of week) after Sub setting the data  (I have it done but not sure
> if it is the best way)


I think what you do works, this might also be a one liner:

SPY[format(index(SPY), "%a") == "Mon", ]

Alternatively

split.default(SPY, format(index(SPY), "%a"))

creates a list of xts objects split by day of the week (Note you need
split.default here because split.xts does something different)

>
> 2.       Rearrange the object (hopefully without destroying the xts object)
> so that my data lines up like a weekly calendar.   So it would look like the
> follow.

Unfortunately, your formatting got all chewed up by the R-help server,
which doesn't like HTML so I'm not quite sure what you want here.

Possibly some black magic like this?

SPY.CL <- Cl(SPY)

length(SPY.CL) <- 7*floor(length(SPY.CL)/7)

dim(SPY.CL) <- c(length(SPY.CL)/7, 7)

But note that this looses time stamps because each row can only have a
single time stamp.

You might also try

to.weekly()

Cheers,

Michael



>
>
>
>
> Long Date Monday
>
> Monday Price
>
> Monday Day Index
>
> Long Date Tuesday
>
> Tuesday Price
>
> Tuesday Day Index
>
> Long Date Wednesday
>
> Wednesday Price
>
> Wednesday Index
>
> Long Date Thursday
>
> Thursday Price
>
> Thursday Index
>
> Friday
>
> Friday Price
>
> Friday Index
>
>
> 1/5/2009
>
> 92.85
>
> 2
>
> 1/6/2009
>
> 93.47
>
> 3
>
> 1/7/2009
>
> 90.67
>
> 4
>
> 1/8/2009
>
> 84.4
>
> 5
>
> 1/9/2009
>
> 89.09
>
> 6
>
>
> 1/12/2009
>
> 86.95
>
> 2
>
> 1/13/2009
>
> 87.11
>
> 3
>
> 1/14/2009
>
> 84.37
>
> 4
>
> 1/15/2009
>
> 91.04
>
> 5
>
> 1/16/2009
>
> 85.06
>
> 6
>
>
> MLK Mondy
>
> MLK Monday
>
> MLK Monday
>
> 1/20/2009
>
> 80.57
>
> 3
>
> 1/21/2009
>
> 84.05
>
> 4
>
> 1/22/2009
>
> 82.75
>
> 5
>
> 1/23/2009
>
> 83.11
>
> 6
>
>
> 1/26/2009
>
> 83.68
>
> 2
>
> 1/27/2009
>
> 84.53
>
> 3
>
> 1/28/2009
>
> 87.39
>
> 4
>
> 1/29/2009
>
> 84.55
>
> 5
>
> 1/30/2009
>
> 82.83
>
> 6
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Thank you,
>
> Douglas
>
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> 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.