CODEWITHSUNDEEP
rloopsplothistogram
sbg
sbg

Reputation: 1772

Loop to create series of graphs from different files

I am trying to plot histograms with long term (several years) mean precipitation (pp) for each day of the month from a series of files. Each file has data collected from a different place (and has a different code). Each of my files looks like this:

 X code year month day  pp  
 1 2867 1945     1   1 0.0  
 2 2867 1945     1   2 0.0   
...

And I am using the following code:

files <- list.files(pattern=".csv")  
par(mfrow=c(4,6))  
for (i in 1:24) {  
    obs <- read.table(files[i],sep=",", header=TRUE)  
    media.dia <- ddply(obs, .(day), summarise, daily.mean<-mean(pp))  
    codigo <- unique(obs$code)  
    hist(daily.mean, main=c("hist per day of month", codigo))  
}

I get 24 histograms with 24 different codes in the title, but instead of 24 DIFFERENT histograms from 24 different locations, I get the same histogram 24 times (with 24 different titles). Can anybody tell me why? Thanks!

Upvotes: 1

Views: 1308

Answers (2)

Andrie
Andrie

Reputation: 179468

There are at least two errors I can see in your code.

  1. There is an error in your ddply statement.
  2. You are passing the wrong variable to hist, thus plotting something that may or may not exist depending on previous session actions.

The problem in your ddply statement is that you are doing an invalid assign (using <- ). Fix this by using =:

media.dia<- ddply(obs, .(day),summarise, daily.mean = mean(pp))

Then edit your hist statement:

hist(media.dia$daily.mean,main=c("hist per day of month",codigo))

I suspect the problem is that you are not passing the correct parameter to hist. The reason that your code actually produces a plot at all, is because in some previous step in your session you must have created a variable called daily.mean (as Brandon points out in the comment.)

Upvotes: 5

James
James

Reputation: 66844

I think the daily.mean calculated in the ddply function is assigned in a separate environment, and does not exist in an environment hist can see.

Try daily.mean<<-mean(pp)

Upvotes: 0

Related Questions