Create adjacency matrix from a list of Id and the corresponding club they are part of

Question

The structure of the data in Stata looks something like this:

id    club_id
1       1
2       1 
3       2
4       2
5       2
6       3
7       3
8       3
9       3

I assume that the network corresponds to the club; hence the adjacency matrix (ordering the data by club) should be block diagonal. I have about 7000 observations. I am new to Mata. I tried to write many variations of the following code but it won't work:

sort id club_id

bys club_id: gen dim = _n     //get the position of the component of the network

mkmat id, mat(id)
mkmat club_id, mat(club_id)
mkmat dim, mat(dim)


local nsubj = _N 

mata:

group_id = st_matrix("id")
AID = st_matrix("club_id")
dim = st_matrix("dim")

adj = J(`nsubj',`nsubj',0)


for (i=1; i<=20; i++) {                   //iteration over each column
    if (club_id[i] == club_id[i+1]) {
        for (k=0, k<= dim[i]-1, k++) {   // the number of columns to be filled
            adj[i, i+k] = 1              // fill the upper part of the matrix
        }
    }

}

end

And I thought to reverse the order of dim to fill the lower part of the matrix. I tried many variations of the code and I don't understand what is wrong.

Answer 1

Here is an approach to create an adjacency matrix without Mata. Maybe that can get you started.

*Replicate data example
clear
set obs 9
gen id = _n 
expand 9
bysort id: gen club_id = _n

*Generate a variable with a value
gen wins = round(runiform()*10)

*Rename the future variable name
gen club = "club"
egen club_id2 = concat(club club_id )
drop club club_id

*Reshape data
reshape wide wins, i(id) j(club_id2) string