[R] Plotting a functional time series

[David Winsemius]

On Feb 24, 2011, at 3:06 PM, Eduardo de Oliveira Horta wrote:

> It seems the code I've sent had typos...
>
> Here's a corrected version:
>
> #################################
> x <- sapply(1:10, function(i)rnorm(1000))
> f <- sapply(1:10, function(i)density(x[,i], from=-5,to=5)$y)
> grid <- density(x[,1], from=-5,to=5)$x
> win.graph()
> persp(grid, 1:10, f,theta=-50, phi=30, d=2)
>
> win.graph()
> opar <- par(mfrow=c(5,2), mar=c(2,2,1,1))
> sapply(1:10, function(i)plot(grid, f[,i], ann=FALSE, type="l"))
> par(opar)
> #################################

This would put them all on one plot:

x <- sapply(1:10, function(i)rnorm(1000))
f <- sapply(1:10, function(i)density(x[,i], from=-5,to=5)$y)
grid <- density(x[,1], from=-5,to=5)$x
pdf()
persp(grid, 1:10, f,theta=-50, phi=30, d=2)
opar <- par(mfrow=c(5,2), mar=c(2,2,1,1))
pdf() ; plot(grid, f[,1], ann=FALSE, type="l") #only one plot
sapply(2:10, function(i) lines(grid, f[,i]))   # the rest with lines()
par(opar); dev.off()

You could color code them with rainbow colors. There is also a lattice  
example with such profiles stacked on top of each other ( in chapter  
14 if I remember correctly) and all the lattice examples are on the  
web at the book site. Search on:

sarkar lattice

>
> Sorry for the mistake.
>
> Best regards,
>
> Eduardo
>
> On Mon, Feb 21, 2011 at 11:18 PM, Eduardo de Oliveira Horta
> <eduardo.oliveirahorta at gmail.com> wrote:
>> Hello,
>>
>> I'm willing to plot a sequence of densities on a 3d graph,  
>> something like
>>
>> -----------------------------------------------------------------
>> x <- sapply(1:10, function(i)rnorm(1000))
>>
>> f <- sapply(1:10, function(i)density(x[,i], from=-5,to=5)$y)
>> grid <- density(x[,1], from=-5,to=5)$x
>>
>> win.graph()
>> persp(grid1, 1:10, f,theta=-50, phi=30, d=2)
>> -----------------------------------------------------------------
>>
>> That is, I simply want to stack the curves
>>
>> -----------------------------------------------------------------
>> win.graph()
>> opar <- par(mfrow=c(5,2), mar=c(2,2,1,1))
>> sapply(1:10, function(i)plot(grid, f[,i], ann=FALSE, type="l"))
>> par(opar)
>> -----------------------------------------------------------------
>>
>> into a functional time series plot (unlike the example above, I'm
>> working with pdf's which are not independent).
>>
>> The result I get using persp() is 'almost' what I want, but I would
>> like to avoid the wireframes connecting the densities along 1:10,
>> because there's no continuity in that direction. In other words, I
>> wanted that the only lines appearing in the plot to be those
>> corresponding to the pdf's f[,1],...f[,10].
>>
>> Thanks once again, and best regards,
>>
>> Eduardo
>>
>>> sessionInfo()
>> R version 2.11.1 (2010-05-31)
>> i386-pc-mingw32
>>
>> locale:
>> [1] LC_COLLATE=Portuguese_Brazil.1252  LC_CTYPE=Portuguese_Brazil. 
>> 1252
>> [3] LC_MONETARY=Portuguese_Brazil.1252 LC_NUMERIC=C
>> [5] LC_TIME=Portuguese_Brazil.1252
>>
>> attached base packages:
>> [1] stats     graphics  grDevices utils     datasets  methods   base
>>
>> other attached packages:
>> [1] rgl_0.92.798     Revobase_4.2.0   RevoScaleR_1.1-1  
>> lattice_0.19-13
>>
>> loaded via a namespace (and not attached):
>> [1] grid_2.11.1       pkgXMLBuilder_1.0 revoIpe_1.0        
>> tools_2.11.1
>> [5] XML_3.1-0
>>
>> ______________________________________________
>> 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.
>>
>
> ______________________________________________
> 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.

David Winsemius, MD
West Hartford, CT