Skip to content

Repeating a series of commands

3 messages · KoopaTrooper, Bert Gunter, Jessica Streicher

KoopaTrooper
# 
I'm trying to figure out how to repeat a series of commands in R and have the
outputs added to a dataframe after each iteration.

My code starts this way...

a<-read.csv("File1.csv")
b<-read.csv("File2.csv")

a$Z<-ifelse(a$Z=="L",sample(1:4,length(a$Z),replace=TRUE),ifelse(a$Z=="M",sample(5:8,length(a$Z),replace=TRUE),ifelse(a$Z=="U",sample(9:10,length(a$Z),replace=TRUE),"")))
a$Z<-as.numeric(a$Z)
b$Z<-ifelse(b$Z=="L",sample(1:4,length(b$Z),replace=TRUE),ifelse(b$Z=="M",sample(5:8,length(b$Z),replace=TRUE),ifelse(b$Z=="U",sample(9:10,length(b$Z),replace=TRUE),"")))
b$Z<-as.numeric(b$Z)

This is basically just starting off with a new and partially random data set
every time that then goes through a bunch of other commands (not shown) and
ends with the following outputs saved.

Output1, Output2, Output3, Output4

where each of these is just a single number. My questions is:

1. How do I repeat the entire series of commands x number of times and save
each of the outputs into a structure like this:
                      Output1      Output2     Output3     Output4
Iteration 1
Iteration 2
Iteration 3
etc.

Not even sure where to start. Are loops the answer? Thanks,





--
View this message in context: http://r.789695.n4.nabble.com/Repeating-a-series-of-commands-tp4645881.html
Sent from the R help mailing list archive at Nabble.com.
Bert Gunter
# 
encapsulate them into a function and call the function ??

-- Bert
On Thu, Oct 11, 2012 at 11:09 AM, KoopaTrooper <ncooper1 at tulane.edu> wrote:

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Jessica Streicher
    

    

      
      #
    

  


  

      
This is pretty confusing writeup, but if you have iterations, then yes, you need a loop or a recursion.

A loop is probably easier, take a look at 

?"for"

or

?apply

for this.

there are several derivates of apply

for example, say your outputs are what the code below gives you for each file, then first make a vector with the files
 filevec<-c("f1.csv","f2.csv")

then you can easily get the "results" (a$z,b$z) as another vector by using

outputs <- sapply(filevec,function(x){
	contents<-read.csv(x);
	something<- ifelse(
		x$Z=="L",sample(1:4,length(x$Z),replace=TRUE),
		ifelse(x$Z=="M",sample(5:8,length(x$Z),replace=TRUE),
		ifelse(x$Z=="U",sample(9:10,length(x$Z),replace=TRUE),"")));
	return(something)
})

and if you want to do this several times:

m<- matrix(ncol=length(filevec),nrow=10)
myResult<- as.data.frame(m)

for(i in 1:10){
	outputs <- sapply(filevec,function(x){
	contents<-read.csv(x);
	something<- ifelse(
		contents$Z=="L",sample(1:4,length(contents$Z),replace=TRUE),
		ifelse(contents$Z=="M",sample(5:8,length(contents$Z),replace=TRUE),
		ifelse(contents$Z=="U",sample(9:10,length(contents$Z),replace=TRUE),"")));
	return(something)
	})
	someStructure[i,]<-outputs
}

If you don't know the number of iterations or outputs beforehand, i'd have a look at "while" and use lists.

Hope this helps.

        
On 11.10.2012, at 20:09, KoopaTrooper wrote: