Re: qplot

[Brandon Hurr]

Do you have an example we could play with? 

I saw this and it might be appropriate... 

http://stackoverflow.com/questions/7962267/cbind-a-df-with-an-empty-df-cbind-fill

On Thu, Sep 20, 2012 at 2:08 PM, SP <heys...@gmail.com> wrote:

I am tyring to use theqplot function in  ggplot2 package and being fairly new to this package and R in general I am a bit stuck. I have five different datasets and every dataset has a column I am interested in plotting. What I would like is to plot these columns together in one graph. But since these columns are of different lengths  I cannot use cbind to create a new  data frame for plotting. Would you have any ideas how to get this working ?

--
You received this message because you are subscribed to the ggplot2 mailing list.
Please provide a reproducible example: https://github.com/hadley/devtools/wiki/Reproducibility
 
To post: email ggp...@googlegroups.com
To unsubscribe: email ggplot2+u...@googlegroups.com
More options: http://groups.google.com/group/ggplot2

[Ben Bond-Lamberty]

Generally speaking, you're going to want to melt() your data, so all
the data are in a single column conditioned by a label column, and
then use qplot(). Note that melt is in the reshape package.

If you want more specific advice, please provide an example, as Brandon said.

Ben

[Brandon Hurr]

Your problem is with ecdf... not with ggplot2. I don't know how to
deal with the output from ecdf...

If you run
> ecdf(z)

You get the same error.

Error in `[.data.frame`(x, order(x, na.last = na.last, decreasing =
decreasing)) :
undefined columns selected



---------- Forwarded message ----------
From: Sarita Paranjpe <heys...@gmail.com>
Date: Fri, Sep 21, 2012 at 2:53 PM
Subject: Re: qplot
To: Brandon Hurr <brando...@gmail.com>


Here you go (z) as my_data


On Fri, Sep 21, 2012 at 3:16 PM, Brandon Hurr <brando...@gmail.com> wrote:
>
> Can you dput(z) so we can see the structure of it and try to plot it?
>
>
> On Fri, Sep 21, 2012 at 1:55 PM, Sarita Paranjpe <heys...@gmail.com> wrote:
>>
>> Hi,
>> I hope this code makes it more easier to interpret my problem :
>> qplot(unique(l), ecdf(l)(unique(l)), geom='step') # with this command I can plot a single column and show the frequency distribution
>> To plot multiple lines (columns of unequal lengths) this is what I did :
>> a1 = as.matrix(a[,2])
>> b1 = as.matrix(b[,2])
>> c1 = as.matrix(c[,2])
>> d1 = as.matrix(d[,2])
>> e1 = as.matrix(e[,2])
>> cbind.fill<-function(...){
>> nm <- list(...)
>> nm<-lapply(nm, as.matrix)
>> n <- max(sapply(nm, nrow))
>> do.call(cbind, lapply(nm, function (x)
>> rbind(x, matrix(, n-nrow(x), ncol(x)))))
>> }
>> bound = cbind.fill(a1,b1,c1,d1,e1) # this rightfully introduced NAs to fill empty rows
>> y = data.frame(bound)
>> z = melt(y) # creates two columns of variable and value
>> qplot(unique(z), ecdf(z)(unique(z)), geom='step') # doesn't work
>>
>> ## Error in `[.data.frame`(x, order(x, na.last = na.last, decreasing = decreasing)) :
>> undefined columns selected
>>
>> ggplot(ecdf(z), aes(x=variable, fill=variable, y=value)) + geom_line() # doesn't work same error as above.
>>
>> Thanks for all your help,
>> SP