The OPTGRAPH Procedure

The following statements use the WEIGHT2= option, and the project groups in a research department as depicted in Figure 1.141. The data set contains the original weight and its inverse, which is used in the calculations of closeness and betweenness.

data LinkSetInDept2;
   input from $1-12 to $13-24 weight weightInv;
   datalines;
Yu          Chapman     3  0.33
Gotti       Chapman     3  0.33
Oliver      Chapman     3  0.33
Snopp       Chapman     3  0.33
Gukrishnan  Leon        1  1
Snopp       Gukrishnan  1  1
Kabutz      Gukrishnan  1  1
Kabutz      Snopp       1  1
Snopp       Leon        1  1
Kabutz      Leon        1  1
Gotti       Oliver      1  1
Gotti       Patrick     1  1
Oliver      Patrick     1  1
Zhuo        Oliver      1  1
Zhuo        Gotti       1  1
Zhuo        Patrick     1  1
Kabutz      Gotti       1  1
Leon        Gotti       1  1
Polark      Yu          2  0.50
Polark      Chang       1  1
Chang       Angel       1  1
Polark      Angel       1  1
Weng        Polark      1  1
Weng        Chang       1  1
Weng        Angel       1  1
Christoph   Yu          2  0.50
Christoph   Nardo       1  1
Christoph   Gotti       1  1
Christoph   Zhuo        1  1
Nardo       Gotti       1  1
Nardo       Zhuo        1  1
Graffe      Yu          2  0.50
Graffe      Hund        1  1
Graffe      Zhuo        1  1
Zhuo        Hund        1  1
;

proc optgraph
   data_nodes   = NodeSetInDept
   data_links   = LinkSetInDept2
   out_nodes    = NodeSetOut;
   performance
      nthreads  = 2;
   centrality
      clustering_coef
      degree    = out
      influence = weight
      close     = weight
      between   = weight
      eigen     = weight
      weight2   = weightInv;
run;

The node data set NodeSetOut now shows the resulting centrality metrics given both weight interpretations.