Optimizing simple for loop in R

Question

I have some lines of code with a for loop that look like this:

somevector2 <- c(length = somevector2_length)

for(string in somevector1){

  df2 <- df1[df1$col1 == string, ]
  ff <- somefunction(df2$col2)
  somevector2 <- c(somevector2, ff)

}

From what i understood initializing the vector with the correct length should make the loop faster, but it still takes quite sometimes although the somefunction(df2$col2) does some simple operations.somevector1 it's just a vector of strings

Is there a way to make this loop faster in R? thank you very much

Edo · Accepted Answer · 2021-01-19 10:04:43Z

2

Sorry, but that's not how you are supposed to post a question on SO. :( You should provide a working example. Also, that's not the way to create a vector of a fixed length.

Let's see a reproducible example of what you posted:

##### this makes your example reproducible

somevector1 <- unique(iris$Species)
df1 <- iris
names(df1) <- paste0("col", 5:1)
somefunction <- sum
somevector2_length <- 3



##### this is your code

# somevector2 <- c(length = somevector2_length) # <- this was wrong
somevector2 <- c()


for(string in somevector1){
 
 df2 <- df1[df1$col1 == string, ]
 ff <- somefunction(df2$col2)
 somevector2 <- c(somevector2, ff)
 
}

So this is the final result:

somevector2
#>  12.3  66.3 101.3

What I suggest you is to use this line of code down here, instead of your code. You will get a similar result (it's a NAMED numeric vector).

tapply(df1$col2, df1$col1, somefunction)
#>    setosa versicolor  virginica 
#>      12.3       66.3      101.3

You can get rid of the names with unname()

answered Jan 19, 2021 at 10:04

Edo

7,8982 gold badges13 silver badges21 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

nonoDa Over a year ago

thanks, i didn't know i had to provide always a reproducible example, i'll make sure to do so in future questions

Edo Over a year ago

not always.. there are some situations were you can't... but generally yes... Also, if this is what you were looking for, I would invite you to accept the answer.

nonoDa Over a year ago

I'm trying to understand this solution you provided, but i don't quite understand the point where this line you provided tapply(df1$col2, df1$col1, somefunction) takes into account the filtering of the df based on the string value provided in the for loop as seen in for(string in somevector1){ df2 <- df1[df1$col1 == string, ].

Edo Over a year ago

that line of code replaces your entire code. If you want just some spefic strings, instead of all of them, just filter your dataframe before that line like this: df1 <- df1[df1$col1 %in% somevector1, ]

Collectives™ on Stack Overflow

Optimizing simple for loop in R

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related