R Programming: Append data in a dataframe within running nested loops

Question

I have following code.

    for(i in seq(from=1, to=8, by=1)) #i is number of stocks from a list
    {
        for(j in seq(from=1, to=8, by=1)) #j is number of trading days from another list
        {
            ## Matching bid and ask prices of each stock for each date and update temp_table
            select_output <- sqldf("select * from temp_table where FLAG == 'new' ")
        }
    }

In current scenario only the last iteration data is staying in select_output which is expected. I want to keep the file open to populate it with data from all the iterations. I don't want to use another loop for inserting data into select_output. Is there any other way possible in R? Please suggest if possible.

Thanks in advance.

Thierry · Accepted Answer · 2016-05-31 09:40:32Z

1

one option is to combine expand.grid, apply and do.call

index <- expand.grid(
  i = 1:8,
  j = 1:8
)
results <- apply(index, 1, function(x){
   x["i"] # holds i
   x["j"] # holds j
   sqldf("select * from temp_table where FLAG == 'new' ")
})
do.call(rbind, results)

answered May 31, 2016 at 9:40

Thierry

18.5k5 gold badges51 silver badges67 bronze badges

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

2 Comments

Zico Over a year ago

no I have more operations inside nested loops. So this way it will be not possible. I am taking data from other dataframes based on i and j positions. I am hoping - if keeping my nested loop intact, you can suggest something on appending data in select_output.

user3603486 Over a year ago

Why can't you do those other operations inside the function within apply, using x["i"] and x["j']?

user3603486 · Accepted Answer · 2016-05-31 10:06:26Z

0

I think this is actually an R weak point. The best thing to do is probably:

results <- data.frame(a=numeric(0), b=character(0), ...) # whatever columns you expect
for(i in 1:8) for(j in s1:7) {
        ## Your code...
        select_output <- sqldf("select * from temp_table where FLAG == 'new' ")
        results <- rbind(results, select_output)      
}

The problem here is that you need in advance to specify the format of results. This can be a good thing - it makes you state explicitly what you expect, and will give an error if those expectations are violated. But it would be nice to have a quick way to bind together a data frame. Maybe the Hadleyverse has a solution?

answered May 31, 2016 at 10:06

user3603486

1 Comment

Thierry Over a year ago

growing objects with rbind() or c() are usually a bad idea

Thierry · Accepted Answer · 2016-05-31 10:52:58Z

0

Here is another solution.

results <- lapply(1:8, function(i){
  #do some stuff with i
  output <- lapply(1:8, function(j){
     #do some stuff with j based on the stuff with i
  })
  do.call(rbind, output)
})
do.call(rbind, results)

answered May 31, 2016 at 10:52

Thierry

18.5k5 gold badges51 silver badges67 bronze badges

Collectives™ on Stack Overflow

R Programming: Append data in a dataframe within running nested loops

3 Answers 3

2 Comments

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related