R help - Dec 2010 - use of 'apply' for 'hist'

Hi all,

##########################################
dof=c(1,2,4,8,16,32)
Q5=matrix(rt(100,dof),100,6,T,dimnames=list(NULL,dof))
par(mfrow=c(2,6))
apply(Q5,2,hist)
myf=function(x){ qqnorm(x);qqline(x) }
apply(Q5,2,myf)
##########################################

These looks ok.
However, I would like to achieve more.

Apart from using a loop,
is there are fast way to 'add' the titles to be more informative?

that is, in the histograms, I want the titles to be 't distribution with
dof=' the degrees of freedom.

I have tried 
apply(Q5,2,hist,xnames=dof)
which does not work;
apply(Q5,2,hist(,xnames=dof));
does not work either

and similarly, how do I add titles to qqnorm plot 
to make them informative?

Thanks!

casper


-- 
View this message in context:
http://r.789695.n4.nabble.com/use-of-apply-for-hist-tp3093811p3093811.html
Sent from the R help mailing list archive at Nabble.com.

Hi,

You can't access the column names from within apply, I'm afraid.
This has been covered previously:
http://www.mail-archive.com/r-help at stat.math.ethz.ch/msg51014.html

That's one of the cases where you actually need to use a for() loop,
though I suppose you could also write something vectorized on the
column names themselves rather than the data frame.

And also, you should check your colnames(Q5) - I don't think they
are what you expect.

Sarah

On Sat, Dec 18, 2010 at 11:49 AM, casperyc <casperyc at hotmail.co.uk>
wrote:
>
> Hi all,
>
> ##########################################
> dof=c(1,2,4,8,16,32)
> Q5=matrix(rt(100,dof),100,6,T,dimnames=list(NULL,dof))
> par(mfrow=c(2,6))
> apply(Q5,2,hist)
> myf=function(x){ qqnorm(x);qqline(x) }
> apply(Q5,2,myf)
> ##########################################
>
> These looks ok.
> However, I would like to achieve more.
>
> Apart from using a loop,
> is there are fast way to 'add' the titles to be more informative?
>
> that is, in the histograms, I want the titles to be 't distribution
with
> dof=' the degrees of freedom.
>
> I have tried
> apply(Q5,2,hist,xnames=dof)
> which does not work;
> apply(Q5,2,hist(,xnames=dof));
> does not work either
>
> and similarly, how do I add titles to qqnorm plot
> to make them informative?
>
> Thanks!
>
> casper
>
>
> --


-- 
Sarah Goslee
http://www.functionaldiversity.org

casperyc wrote:
> 
> Hi all,
> 
> ##########################################
> dof=c(1,2,4,8,16,32)
> Q5=matrix(rt(100,dof),100,6,T,dimnames=list(NULL,dof))
> par(mfrow=c(2,6))
> apply(Q5,2,hist)
> myf=function(x){ qqnorm(x);qqline(x) }
> apply(Q5,2,myf)
> ##########################################
> 
> Apart from using a loop, is there are fast way to 'add' the titles
to be
> more informative?
> 
> 
If you type "apply" (no ()) you will find that it is not much more
than an
decorated generic loop. I believe that this is case where using a loop.
Otherwise, you could convert you matrix to a data frame, add the degrees of
freedom.... too much work for a loop where plotting anyway has the largest
overhead.

Others may disagree, but I believe the apply-wars have ended at the time
when browser wars ended.

Dieter

-- 
View this message in context:
http://r.789695.n4.nabble.com/use-of-apply-for-hist-tp3093811p3093847.html
Sent from the R help mailing list archive at Nabble.com.

On Sat, Dec 18, 2010 at 11:49 AM, casperyc <casperyc at hotmail.co.uk>
wrote:
>
> Hi all,
>
> ##########################################
> dof=c(1,2,4,8,16,32)
> Q5=matrix(rt(100,dof),100,6,T,dimnames=list(NULL,dof))
> par(mfrow=c(2,6))
> apply(Q5,2,hist)
> myf=function(x){ qqnorm(x);qqline(x) }
> apply(Q5,2,myf)
> ##########################################
>
> These looks ok.
> However, I would like to achieve more.
>
> Apart from using a loop,
> is there are fast way to 'add' the titles to be more informative?
>
> that is, in the histograms, I want the titles to be 't distribution
with
> dof=' the degrees of freedom.
>
> I have tried
> apply(Q5,2,hist,xnames=dof)
> which does not work;
> apply(Q5,2,hist(,xnames=dof));
> does not work either
>
> and similarly, how do I add titles to qqnorm plot
> to make them informative?
Loop over the column headings rather than over the data itself.  Be
sure that dof has class "character".  Always include set.seed if you
post random numbers to r-help so the results are reproducible.  Always
change the par() back after you are finished.  A number of other
stylistic improvements are shown below as well.  Change main= and
xlab= as you like.

	set.seed(123)
	dof <- as.character(c(1,2,4,8,16,32))
	Q5 <- matrix(rt(100, dof), 100, 6, T, dimnames = list(NULL, dof))
	opar <- par(mfrow = c(2, 6))
	sapply(dof, function(x) hist(Q5[, x], main = x, xlab = ""))
	myf <- function(x) { qqnorm(Q5[, x], main = x, xlab = "");
qqline(Q5[, x]) }
	sapply(dof, myf)
	par(opar)


-- 
Statistics & Software Consulting
GKX Group, GKX Associates Inc.
tel: 1-877-GKX-GROUP
email: ggrothendieck at gmail.com