|
Notes:
UPDATES GraphVar beta 0.4
1. Create Connectivity Matrix:
- Acceleration of connectivity matrix generation (achieved through
vectorization)
2. Network Construction:
- Calculate and export graph topological measures now also without
applying thresholds (i.e., on the fully connected matrix)
- Modify weights in the matrices by using "absolute weights" or
"negative weights to zero"
3. Network Calculations
- GraphVar will warn you at the densities were negative weights
will enter your calculations
- Inclusion of "Rich Club Coefficient" (please ignore the node
labels in the export -> each column stands for level k)
4. Statistics
- Inclusion of non-parametric testing on the metric level for
correlational analysis (i.e., testing against shuffled data)
5. Results Viewer
- plot single subject association matrices (e.g. for QC)
- plot mean connectivity matrices (i.e., connectivity across
subjects) to determine the direction of your effects in more detail
6. Network Inspector
- open your identified graph component immediately in
BrainNetViewer for 3D visualization (BrainNetViewer must be added
to Matlab with subfolders)
Changes:
|
