Dynamically create a variable within a loop and add to dataframe-CodePudding

In STATA creating a variable dynamically within a loop is easy because the quotations `' identify the iterator. This example is to create a binary variable Y200X that takes value 1 if the Year is lesser than year 200X:

set obs 10
gen Year = 2005
replace Year = 2010 if _n > 4

forvalues y = 2005(1)2020 {
    gen byte Y`y' = 0
    replace Y`y' = 1 if Year < `y' 
    }

In R the iterator cannot be used directly for creating the variable name. The best I found was first create variables in the loop then assemble them back into the dataframe outside the loop:

Year <- c(2005,2010,1996,1994,2001,2006,2019,2021, 2018,1987)
ls.output <- as.data.frame(Year)

for(y in 2005:2020) {
  assign(paste0("Y",y), ifelse(ls.output$Year < y, 1, 0))
}
ls.output<- cbind(ls.output, Y2005,Y2006,Y2007,Y2009, Y2010)

Is there a better way to do this directly in the loop?

CodePudding user response：

Column names can be pasted together. Skip the step of creating separate variables and then adding them to the data frame as columns, instead add them directly:

for(y in 2005:2020) {
  ls.output[, paste0("Y", y)] <- ifelse(ls.output$Year < y, 1, 0)
}

ls.output
#    Year Y2005 Y2006 Y2007 Y2008 Y2009 Y2010 Y2011 Y2012 Y2013 Y2014 Y2015 Y2016 Y2017 Y2018 Y2019 Y2020
# 1  2005     0     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1
# 2  2010     0     0     0     0     0     0     1     1     1     1     1     1     1     1     1     1
# 3  1996     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1
# 4  1994     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1
# 5  2001     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1
# 6  2006     0     0     1     1     1     1     1     1     1     1     1     1     1     1     1     1
# 7  2019     0     0     0     0     0     0     0     0     0     0     0     0     0     0     0     1
# 8  2021     0     0     0     0     0     0     0     0     0     0     0     0     0     0     0     0
# 9  2018     0     0     0     0     0     0     0     0     0     0     0     0     0     0     1     1
# 10 1987     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1     1

CodePudding user response：

Use outer where Year is as defined in the question.

data.frame(Year,  outer(Year, setNames(2005:2010, paste0("Y", 2005:2010)), `<`))

giving:

   Year Y2005 Y2006 Y2007 Y2008 Y2009 Y2010
1  2005     0     1     1     1     1     1
2  2010     0     0     0     0     0     0
3  1996     1     1     1     1     1     1
4  1994     1     1     1     1     1     1
5  2001     1     1     1     1     1     1
6  2006     0     0     1     1     1     1
7  2019     0     0     0     0     0     0
8  2021     0     0     0     0     0     0
9  2018     0     0     0     0     0     0
10 1987     1     1     1     1     1     1