subset dataframes based on rows values and create list of new data frames in R-CodePudding

I have a very large data frame with a column with years.

Say this df looks like this:

df <- data.frame(var1 = c("A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M", "N"),
                 ANO = c(1990, 1994, 2008, 2019, 1992, 2003, 2008, 2001, 2015, 1998, 2012, 1985, 1970, 1965))

I want to subset this large df into a list of dfs based on the column 'ANO' with a 'smaller or equal than' expression.

Importantly: The original dataset has values ranging c.1953 -2020, but I only need the new 31 dfs for the years 1990-2020. However, the individual dfs need to have the rows for the years before 1990 as well. This is, the new list would need to have the 31 dfs (df_1990 - df_2020) and, for instance, the df for 1990 would need to include all rows with year before 1990; df_1991, all rows with years before 1991, and so on until 2020.

It works if I specify each dataframe I want to subset individually, like this:

new_list  <- list()
new_list$df_2010 <- df[df$ANO <= 2010,]
new_list$df_2011 <- df[df$ANO <= 2011,]
new_list$df_2012 <- df[df$ANO <= 2012,]

etc.

I would like to avoid copying 30 lines every time, so I tried something like this:

years <- seq(from=1990, to=2020)

new_list <- list()

for(i in 1:length(years)) {
  
  df_loop <- df[df$ANO <= years[i],]
  
  new_list <- df_loop[[i]]
  
  }

But it does not work, I get this error:

Error in .subset2(x, i, exact = exact) : subscript out of bounds

Any suggestions on how to achieve that result?

CodePudding user response：

You can use lapply to iterate over every value in df$ANO, and use subset to pick rows that is smaller than or equal to ANO. Wrap the lapply function with setNames to name the sub-list data frames.

Edit: If you want to only include a specific range, you can play around with the X argument in lapply. Here we'll only loop through 1990 to 2020 in lapply.

new_list <- setNames(lapply(sort(df$ANO)[sort(df$ANO) >= 1990], function(x) subset(df, ANO <= x)), paste0("df_", sort(df$ANO)[sort(df$ANO)>=1990]))

Output

new_list

$df_1990
  var1  ANO
1    A 1990

$df_1992
  var1  ANO
1    A 1990
5    E 1992

$df_1994
  var1  ANO
1    A 1990
2    B 1994
5    E 1992

$df_1998
   var1  ANO
1     A 1990
2     B 1994
5     E 1992
10    J 1998

$df_2001
   var1  ANO
1     A 1990
2     B 1994
5     E 1992
8     H 2001
10    J 1998

$df_2003
   var1  ANO
1     A 1990
2     B 1994
5     E 1992
6     F 2003
8     H 2001
10    J 1998

$df_2008
   var1  ANO
1     A 1990
2     B 1994
3     C 2008
5     E 1992
6     F 2003
7     G 2008
8     H 2001
10    J 1998

$df_2008
   var1  ANO
1     A 1990
2     B 1994
3     C 2008
5     E 1992
6     F 2003
7     G 2008
8     H 2001
10    J 1998

$df_2012
   var1  ANO
1     A 1990
2     B 1994
3     C 2008
5     E 1992
6     F 2003
7     G 2008
8     H 2001
10    J 1998
11    K 2012

$df_2015
   var1  ANO
1     A 1990
2     B 1994
3     C 2008
5     E 1992
6     F 2003
7     G 2008
8     H 2001
9     I 2015
10    J 1998
11    K 2012

$df_2019
   var1  ANO
1     A 1990
2     B 1994
3     C 2008
4     D 2019
5     E 1992
6     F 2003
7     G 2008
8     H 2001
9     I 2015
10    J 1998
11    K 2012