[R] a more elegant way to get percentages?

[Monica Pisica]

Hi everybody,


I am amazed how quick i got my answer ;-) I have to recognize that Gabor's code really puts to shame my skills in doing any programming in R. Is there any book or documentation which really explains in details all these neat tricks from {stats} like ave (i even didn't know this function existed), apply and all its friends (sapply, tapply, etc) ? To be honest it took me quite a while to come up with the "fancy" subscripting to get my persantages ;-))

thank you so much, i really appreciate your help,

Monica

> Date: Thu, 13 Mar 2008 09:45:05 -0400
> From: ggrothendieck at gmail.com
> To: pisicandru at hotmail.com
> Subject: Re: [R] a more elegant way to get percentages?
> CC: r-help at r-project.org
>
> Assuming your x is as follows:
>
> x <- data.frame(locat = c("a", "b", "b", "c", "c", "c", "c", "d", "d", "d"),
> val = c(5, 5, 15, 5, 20, 5, 10, 5, 15, 10))
>
> Try this:
>
> x$percent1 <- ave(x$val, x$locat, FUN = function(x) 100*x/sum(x))
>
> On Thu, Mar 13, 2008 at 9:36 AM, Monica Pisica  wrote:
>>
>> Hi,
>>
>> I am trying to get percentages in a more elegant way. I have a data.frame with locations and values (counts) of species at that location. Each location is repeated for each species i have values for and i would like to get percentages of each species at that location. I am not sure if i am clear in my explanations so i will paste my code below:
>>
>> #####################
>>
>>> x
>> locat val
>> 1 a 5
>> 2 b 5
>> 3 b 15
>> 4 c 5
>> 5 c 20
>> 6 c 5
>> 7 c 10
>> 8 d 5
>> 9 d 15
>> 10 d 10
>>> loc1 <- x$locat
>>> n <- length(loc1)
>>> locuniq1 <- unique(loc1)
>>> m <- length(locuniq1)
>>> counts <- seq(1:m)
>>>
>>> for (i in 1:m) {
>> + count <- 0
>> + for (j in 1:n) {
>> + if (loc1[j]==locuniq1[i]) count <- count+1
>> + counts[i] <- count
>> + }
>> + }
>>>
>>> percent1 <- rep(0,n)
>>> j <- 0
>>> for (i in 1:m) {
>> +
>> + b <- x[(j+1):(j+counts[i]),]
>> + total <- sum(b$val)
>> + percent1[(j+1):(j+counts[i])] <- round(apply(as.matrix(b$val), 1, function(x) {x*100/total}),2)
>> + j = j+counts[i]
>> + }
>>> x1 <- cbind(x, percent1) # this is the result i want
>>> x1
>> locat val percent1
>> 1 a 5 100.00
>> 2 b 5 25.00
>> 3 b 15 75.00
>> 4 c 5 12.50
>> 5 c 20 50.00
>> 6 c 5 12.50
>> 7 c 10 25.00
>> 8 d 5 16.67
>> 9 d 15 50.00
>> 10 d 10 33.33
>>>
>> ################
>>
>> I am wondering if there is any way to do it more efficiently, much more that the first loop which gives how many times each location is present in the data.frame is slow if you have a larger data.frame and not only 10 rows.
>>
>> Thanks for any input and sorry if the email is on the long side,
>>
>> Monica
>>
>>
>> _________________________________________________________________
>> [[elided Hotmail spam]]
>>
>> ______________________________________________
>> 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.
>>

_________________________________________________________________


08