Question

I apologize if this question seems rudimentary. I’m a bit unclear on if it’s possible / how to use for loops with a list of data frames in R.

I have dataframes that are subsets of a larger dataframe:

billysuzy <- subset(userlist,user1 %in% c("billy","suzy"))
joefrank<- subset(userlist,user1 %in% c("joe","frank"))
georgelenny <- subset(userlist,user1 %in% c("george","lenny"))

I would like to loop through and find correlation values for the same variables (time and simscore) for each subset.

cor(time, simscore)

However, I am quite uncertain how to structure such a for loop and unsure how to structure a function to attach and detach within lapply.

I was thinking:

somelist <- list(billysuzy, joefrank, georgelenny)

with lapply:

corz <-  lapply(somelist, function(df)  detach(), attach(df),cor(time, simscore))
lapply(corz, print)    

or with for:

for (i in 1:length(somelist)){
      detach()
     attach(somelist[i])
    cor(time, simscore)
    }

I apologize for offending anyone’s R sensibilities or if the seems like two questions – the answer is any which describes how to do the one thing i.e. loop through the dfs in the list and find there corr values.

Answer 1

I would really recommend avoiding attach and detach.

You could try something like this

corz <-  lapply(somelist, function(df)  cor_time_sim = cor(df$time, df$simscore))

Or

corz <-  lapply(somelist, function(DF)  {with(DF, corst <-  cor(time, simscore))})

Or

corz <-  lapply(somelist, with, corst <-  cor(time, simscore))