Subset dataframe using a loop

Question

I have a data frame that looks like this:

index   ID   date              Amount
2       1001 2010-06-08         0
21      1001 2010-10-08        10
6       1002 2010-08-16        30
5       1002 2010-11-25        20
9       1003 2010-01-01         0
8       1003 2011-03-06        10
12      1004 2012-03-12        10
11      1004 2012-06-21        10
15      1005 2010-01-01        30
13      1005 2010-04-06        20

I want to subset this data so that I have new data frames, one for each ID like this

index   ID   date              Amount
2       1001 2010-06-08         0
21      1001 2010-10-08        10

And

6       1002 2010-08-16        30
5       1002 2010-11-25        20

And so on.

I don't need to save the new data frames, but use it to perform some basic calculations. Also I want to do this on my entire table consisting of more than 10000 IDs and hence the need for a loop. I tried this

    temp <- data.frame(Numb=c(),Dt=c(),Amt=c())
for (i in seq_along(stNew$ID)){
   temp[i,] <- subset(stNew, stNew[i,]==stNew$ID[i])
}

But that didn't work. Any suggestions?

Hi and welcome to SO! My spontaneous suggestion is that you should try to search SO (and elsewhere) for an answer. To perform something per group in a dataframe is one of the most commonly asked questions on SO, and you will surely find some nice answers you can adapt to your own data. [This](http://stackoverflow.com/questions/3505701/r-grouping-functions-sapply-vs-lapply-vs-apply-vs-tapply-vs-by-vs-aggrega) and [this](http://lamages.blogspot.se/2012/01/say-it-in-r-with-by-apply-and-friends.html) may get you started. Cheers. — Henrik, Nov 09 '13 at 20:00
Henrik - thanks. As a matter of fact i did search through and found a couple which were extremely useful. Thanks for your links as well. — Bala Deshpande, Nov 10 '13 at 22:42
Great! Thus, no need for splitting or subsetting your data frame. — Henrik, Nov 10 '13 at 22:45

score 13 · Answer 1 · edited Aug 06 '20 at 18:58

Take a look at the list2env and split function. Hereby some examples using the iris dataset.

Two way:

list_df <- split(iris, iris$Species) #split the dataset into a list of datasets based on the value of iris$Species
list2env(list_DF, envir= .GlobalEnv) #split the list into separate datasets

One way:

list2env(split(iris, iris$Species), envir = .GlobalEnv)

Or you can assign custom names for the new datasets with a for loop:

iris_split <- split(iris, iris$Species)
new_names <- c("one", "two", "three")
for (i in 1:length(iris_split)) {
  assign(new_names[i], iris_split[[i]])
}

Updates with examples

score 8 · Answer 2 · answered Nov 09 '13 at 19:32

8

may be like this

    IDs<-unique(df$ID)
    for (i in 1:length(IDs)){ 
    temp <- df[df$ID==IDs[i],]
    #more things to do with temp
    }

answered Nov 09 '13 at 19:32

Ananta

3,671
3
22
26

score 6 · Answer 3 · answered Jun 27 '16 at 05:00

6

    iris_split <- split(iris, iris$Species)

Dynamically you can assign the data.frame name

    new_names <- as.character(unique(iris$Species))

    for (i in 1:length(iris_split)) {
    assign(new_names[i], iris_split[[i]])
    }

answered Jun 27 '16 at 05:00

B G Manjunath Prasad

78
1
5

Subset dataframe using a loop

3 Answers3

Linked