search

bmkramer / 101innovations-survey-data

Stringing beads - identifying research workflows from tool usage data (clustering)
11 stars 5 forks source link

Network visualization / network analysis #11

Open bmkramer opened 8 years ago

bmkramer commented 8 years ago

Network analysis of tool usage based on co-occurrence

Relation with #5 Relation with #12

RMHogervorst commented 8 years ago

I'm not familiar with these techniques, do you have any materials to start wiht?

bmkramer commented 8 years ago

I hope to start with the adjacency matrix of co-occurrence frequencies and load that into Gephi as per here: https://github.com/gephi/gephi/issues/1143.

This will require some permutations that would be great if coded into R (but I'll probably do them in Calc for now):

  1. replace column names in first row with sequential numbers (as in first column), remove last column (with names)
  2. create separate csv with two columns: sequential numbers and names (e.g. current first and last column
    • The first permutation is necessary to import the csv into Gephi and have it processed properly into an edge table
    • The second permutation enables import in Gephi as node table.

I'll keep a log in this comment thread on my workflow.

bmkramer commented 8 years ago

I created a separate branch for this and will document my workflow there in README.md