How to Make Function For Multiple Datasets in R like macro in SAS

Question

I have data from Jan 2021 - Feb 2022 (the names data like CARD_202101, CARD_202102, CARD_202103 until CARD_202202) each data contain variables:

• CIF
• Date
• Descriptions

How can I make function in R. so I just can have CIF and Date variables for all datasets from Jan 2021 - Feb2022

example:

CARD_202101 <- data.frame(CIF = c(1,2,3), Date = c('2021-01-01', '2021-01- 
               02','2021-01-01'), Descriptions = c("a", "b", "c"))
CARD_202102 <- data.frame(CIF = c(1,6,3), Date = c('2021-02-01', '2021-02- 
               02','2021-01-01'), Descriptions = c("a", "b", "c"))
....


CARD_202202 <- data.frame(CIF = c(4,2,3), Date = c('2022-02-01', '2022-02- 
               02','2022-02-01'), Descriptions = c("a", "b", "c"))

I just want each dataset just only contain CIF and Date Variables, like

CARD_202101 <- data.frame(CIF = c(1,2,3), Date = c('2021-01-01', '2021-01- 
               02','2021-01-01'))

CARD_202102 <- data.frame(CIF = c(1,6,3), Date = c('2021-02-01', '2021-02- 
               02','2021-01-01'))
....


CARD_202202 <- data.frame(CIF = c(4,2,3), Date = c('2022-02-01', '2022-02- 
               02','2022-02-01'))

I need looping through all dataset

Answer 1

I will asume few things first:

1. naming convention of data is like: it starts with CARD_ and followed by 6 digits
2. I can use data.table package
3. objects are in global environment

If so i recomend loop through all the object that meet naming convention and bind them using data.table::rbindlist like this:

bind_datasets <- function()
{
    data.table::rbindlist(
        l = lapply(
            X = ls(envir = globalenv(), pattern = "^(CARD_)\\d{6}$"),
            FUN = function(i)
            {
                res <- get(x = i, envir = globalenv())
                res <- subset(x = res, select = c("CIF", "Date"))
                return(res)
            }
        )
    )    
}

This function:

1. search global environment for object named with established pattern
2. for each object retrieves CIF and Date column
3. bind object one after another.

EDIT:

After your comment I think the answer is the code below:

# find all the data in global environment which comply with a pattern
datasets <- ls(envir = globalenv(), pattern = "^(CARD_)\\d{6}$")
# loop through it and each time assign to them their subset
for (dst in datasets)
{
    res <- get(x = dst, envir = globalenv())
    assign(
        x = dst, 
        value = subset(x = res, select = c("CIF", "Date")), 
        envir = globalenv()
    )
}