R help - Jun 2010 - building time series/zoo/its from a data frame

Dear R People:

I have the following data frame:
> x.df
       date        cond freq
1  04/01/09       Fever   12
2  04/02/09       Fever   11
3  04/03/09       Fever   10
4  04/04/09       Fever   13
5  04/05/09       Fever    6
6  04/01/09        Rash    6
7  04/02/09        Rash   10
8  04/03/09        Rash    9
9  04/04/09        Rash   10
10 04/05/09        Rash    8
11 04/01/09 Respiratory   12
12 04/02/09 Respiratory    9
13 04/03/09 Respiratory    6
14 04/04/09 Respiratory   11
15 04/05/09 Respiratory   11
>
I would like to generate 3 time series (or zoo objects or its); one
for Fever, one for Rash, and one for Respiratory.   There are 2
questions here, please:
a. How do I generate the series, please?
b. Which is best, time series, zoo objects, or its objects, please?

Thanks,
Erin


-- 
Erin Hodgess
Associate Professor
Department of Computer and Mathematical Sciences
University of Houston - Downtown
mailto: erinm.hodgess at gmail.com

Hi:

Here's one approach, converting the sub-data frames to zoo objects:

# Function to convert one of the data frames to a zoo object
makezoo <- function(df) {
     require(zoo)
     date <- as.Date(df[, 1], format = '%m/%d/%y')
     with(df, zoo(freq, date))
    }
library(zoo)

# Split the data frame by condition:
tt <- split(x.df[, -2], x.df$cond)

# > str(tt[[1]])
# 'data.frame':   5 obs. of  2 variables:
#  $ date: Factor w/ 5 levels "04/01/09","04/02/09",..: 1 2
3 4 5
#  $ freq: int  12 11 10 13 6

# Apply the makezoo function to each component of the list tt:
(tser <- lapply(tt, makezoo))

$Fever
04/01/09 04/02/09 04/03/09 04/04/09 04/05/09
      12       11       10       13        6

$Rash
04/01/09 04/02/09 04/03/09 04/04/09 04/05/09
       6       10        9       10        8

$Respiratory
04/01/09 04/02/09 04/03/09 04/04/09 04/05/09
      12        9        6       11       11

# Check that the classes are OK:
> sapply(tser, class)
      Fever        Rash Respiratory
      "zoo"       "zoo"       "zoo"

You can then do things like plot.ts, for example:

lapply(tser, plot.ts)

HTH,
Dennis

On Wed, Jun 2, 2010 at 8:37 AM, Erin Hodgess
<erinm.hodgess@gmail.com>wrote:
> Dear R People:
>
> I have the following data frame:
>
> > x.df
>       date        cond freq
> 1  04/01/09       Fever   12
> 2  04/02/09       Fever   11
> 3  04/03/09       Fever   10
> 4  04/04/09       Fever   13
> 5  04/05/09       Fever    6
> 6  04/01/09        Rash    6
> 7  04/02/09        Rash   10
> 8  04/03/09        Rash    9
> 9  04/04/09        Rash   10
> 10 04/05/09        Rash    8
> 11 04/01/09 Respiratory   12
> 12 04/02/09 Respiratory    9
> 13 04/03/09 Respiratory    6
> 14 04/04/09 Respiratory   11
> 15 04/05/09 Respiratory   11
> >
>
> I would like to generate 3 time series (or zoo objects or its); one
> for Fever, one for Rash, and one for Respiratory.   There are 2
> questions here, please:
> a. How do I generate the series, please?
> b. Which is best, time series, zoo objects, or its objects, please?
>
> Thanks,
> Erin
>
>
> --
> Erin Hodgess
> Associate Professor
> Department of Computer and Mathematical Sciences
> University of Houston - Downtown
> mailto: erinm.hodgess@gmail.com
>
> ______________________________________________
> 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.
>
	[[alternative HTML version deleted]]

r-help seems to have rejected the first attempt at sending this so I
am trying again.

read.zoo has a split= argument which lets you read in the data and
split it all at the same time.  You can also use it on a data frame.
Due to a bug which is fixed in the development version, you will need
the development version of zoo which you can access as shown below.
First we show reading it in using read.zoo and splitting it at the
same time.  Next we assume its already been read in as a data frame in
which case we can also use read.zoo on a data.frame:

library(zoo)
source("http://r-forge.r-project.org/scm/viewvc.php/*checkout*/pkg/zoo/R/read.zoo.R?revision=717&root=zoo")

Lines <- "date        cond freq
1  04/01/09       Fever   12
2  04/02/09       Fever   11
3  04/03/09       Fever   10
4  04/04/09       Fever   13
5  04/05/09       Fever    6
6  04/01/09        Rash    6
7  04/02/09        Rash   10
8  04/03/09        Rash    9
9  04/04/09        Rash   10
10 04/05/09        Rash    8
11 04/01/09 Respiratory   12
12 04/02/09 Respiratory    9
13 04/03/09 Respiratory    6
14 04/04/09 Respiratory   11
15 04/05/09 Respiratory   11"

# 1. reading it in and splitting at the same time
z <- read.zoo(textConnection(Lines), header = TRUE, split = "cond",
format = "%m/%d/%y")

# or, 2. if its already a data frame such as this one we can still use read.zoo
DF <- read.table(textConnection(Lines), header = TRUE)
z2 <- read.zoo(DF, split = 2, format = "%m/%d/%y")