0

Following up on my previous question, I would like to create degree centrality for with the following data.

I tried to replicate the previous example on stackoverflow (how to find degree centrality of nodes in a matrix?) but I had some challenges.

 a <- c('nancy','bill','bob','badri','bill','kiron','david')
 b <- c('martial-arts','dance','sports','judo','judo','judo','judo')

 df <- data.frame(a, b)

From this data i need to create an adjacency matrix : (where u1, u2, u3, u4,u5,u6 represents nancy, bill,bob, badri users and how they are connected through the clubs they are participating in). Can someone help?

I am having difficulty in creating this adjacency matrix from raw data. Can someone suggest any ideas on how to proceed.

Community
  • 1
  • 1
user3570187
  • 1,743
  • 3
  • 17
  • 34
  • I am trying to create an adjacency matrix and i have unequal rows and columns(not a square one). My users are much more than groups. Any suggestions on how to proceed? Thx guys! – user3570187 Sep 03 '16 at 20:24

1 Answers1

3

This solution comes via Solomon Messing:

M <- as.matrix(table(df))
M <- tcrossprod(M)
diag(M) <- 1
M
#        a
# a       badri bill bob david kiron nancy
#   badri     1    1   0     1     1     0
#   bill      1    1   0     1     1     0
#   bob       0    0   1     0     0     0
#   david     1    1   0     1     1     0
#   kiron     1    1   0     1     1     0
#   nancy     0    0   0     0     0     1

See the complete post for guidance on how to do this in a sparse matrix.

Weihuang Wong
  • 12,868
  • 2
  • 27
  • 48