Repeat same action with function and apply familly

Question

I'm new on R (and I use R-studio) and I have to analyze a big data frame (60 variables for 10 000 observations). My data frame had a column name specie with lot of different animals species in there. The goal of my work it's to have results of 8 differents species, so I have to work on there separately. I start with building different subset (like I learn in school) and with awesome packages(special thanks to dplyr & tdyr). But now I have to repeat many identical (or nearly identical) actions on each of the 8 species, so I spent much time to copy/paste and when I make a mistake I must verify and change mistakes on thousands of lines. Then I try to learn about loops et apply family functions. But I can't do something good.

There is an exemple of an action I do on a specie with the traditional way (organize data):

espece_td_a <- subset(BDD, BDD$espece == "espece A" & BDD$placette =="TOTAL")%>%
  select(code_site,passage,adulte)%>% 
  spread(passage, adulte)              
  espece_td_a <- full_join(B.irene_td_a, BDD_P3_TOT_site)
  espece_td_a <- replace(espece_td_a, is.na(espece_td_a),0)  
  espece_td_a$P1[B.irene_td_a$P1>0]<-1                       
  espece_td_a$P2[B.irene_td_a$P2>0]<-1
  espece_td_a$P3[B.irene_td_a$P3>0]<-1
  write.csv(espece_td_a, file = "espece_td_a.csv")

BDD is my data frame.

BDD_P3_TOT_site is vector (or data frame with 1 columns and many rows ?) built with BDD

This "traditional way" work for me, but I must do something like that so many times! And it takes a lot of time...

Then I tried to "apply" this with function :

f <- function(x)
{
  select(code_site, passage, adulte)%>%
    spread(x, x$passage, x$adulte)%>%
    full_join(x, BDD_P3_TOT_site) -> x
  x <- replace(x, is.na(x),0)
  x$P1[x$P1>0]<-1
  x$P2[x$P2>0]<-1
  x$P3[x$P3>0]<-1
}

I wish apply this function to my dataset with lapply (with my 8 species in list):

l <- c("espece_a","espece_b","espece_c")

lapply(l,f(x))

Problems :

I know that is a wrong formulation for lapply if I want take my species into BDD.
the function doesn't want work: I already made 8 subsets (for each of my interest species) In my global environment: espece_a; espece_b...

Then I wanted to put my subset one by one into my function:

> f(espece_a)

Error in select_(.data, .dots = lazyeval::lazy_dots(...)) : Show Traceback object 'code_site' not found Rerun with Debug

I wish that my table appears in my Globlal env with a name that make me able to recognize it (ex: "espece_td_a")

Geia sou Theodore :). Welcome to SO. Please go through [this link](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) and create a sample data set for us to play with. Also, add your expected output. — Sotos, Jul 04 '17 at 16:00
Ok, after a first read of your post, It seems like you need to study `dplyr` syntax. Quite a few things wrong there. Also in order to make functions using `dplyr` you need to read on standard / Non-Standard Evaluation. Also, I think you are really over complicating things. If you provide a sample we should be able to help you — Sotos, Jul 04 '17 at 16:09

score 0 · Answer 1 · answered Jul 04 '17 at 15:38

0

You have 3 issues relating to your use of lapply:

You need to return the object x at the end of the f function:
l should be a list of dataframes not just a vector of dataframe names, i.e. l <- list(espece_a,espece_b,espece_c)
When using lapply with an existing function, you only need to pass the name of the function, i.e. lapply(l,f)

Hopefully this should solve your problem.

answered Jul 04 '17 at 15:38

LucyMLi

657
4
14

I solve the function problem, if I apply the function on one of my subset, it change it. But when I use lapply I just have a list with the name of my subset, I want all my subset in the list change by my function. – Théodorus Kyritsos Jul 05 '17 at 12:09

score 0 · Answer 2 · answered Jul 05 '17 at 12:14

I solve the function problem :

f <- function(X){
  X <- select(X, code_site, passage, adulte)%>%
   spread(passage, adulte) 
  X <- full_join(X, BDD_P3_TOT_site)
  X <- replace(X, is.na(X),0)
  X$P1[X$P1>0]<-1
  X$P2[X$P2>0]<-1
  X$P3[X$P3>0]<-1
  X <- return(X)
    }

test <- f(espece_a)

Repeat same action with function and apply familly

2 Answers2