Question

I created a list which contains 105 matrices as follows:

m<-vector("list",105)
for (i in 2:105) {
    m[[i-1]]<-Datos[(x[i-1]+1):x[i],1:14] }
m[[105]]<-Datos[(x[105]+1):533195,1:14]

For example a part of my matrix number 104 returns (In columns):

m[[104]]

    ID:                    
     8866 
     8866 
     8866 
     8866 
     8866 
     8866 
     8866 
     8866

  Date: 
     1990-4-15 
     1990-4-16
     1990-4-17
     1990-4-18
     1990-4-15
     1990-4-16
     1990-4-17
     1990-4-18

    Series: 
        APV 
        APV 
        APV 
        APV 
        INV 
        INV 
        INV 
        INV

These are some of my columns of the matrix. What I would like is to split this matrix using the series columns. I think it would be like a list of a list depending of the number of different Series there are. In this case there are 2: APV and INV ( Note that I don’t know the names of the series for each matrix, so there must be a function that extract the unique different series)

In summary, I would like that:

m[[104]][[1]] returns:

        ID:                    
         8866 
         8866 
         8866 
         8866 


      Date: 
         1990-4-15 
         1990-4-16
         1990-4-17
         1990-4-18


        Series: 
            APV 
            APV 
            APV 
            APV 



And m[[104]][[2]] returns:

        ID:                    
         8866 
         8866 
         8866 
         8866 


      Date: 
         1990-4-15 
         1990-4-16
         1990-4-17
         1990-4-18


        Series: 

            INV 
            INV 
            INV 
            INV

Or maybe you come up with a more efficient way to do this.

PD: Didn’t know how to put the columns at the same level

Answer 1

Try this:

library(plyr)
lapply(m, dlply, "Series")

or:

lapply(m , function(x)split(x, x$Series))