Decision Tree Demo
This demo shows how to create an interactive decision tree from a graph.
The decision tree contains different types of nodes.
-
Orange nodes are the current options that can be clicked in order to expand the next
tree level.
- Blue nodes are nodes that have been previously chosen.
- Green nodes represent options along the path that have not been chosen.
- Red nodes are end nodes.
Things to Try
Decision Tree
- Browse the Sample graphs.
-
Click the orange nodes
to make a decision and expand the next level of the tree.
-
Click the blue or green nodes
to reset the decision tree back to this level and select the clicked option.
- Click Restart to reset the current decision tree.
-
Click the edit button
to edit the current sample and view the complete graph.
Edit View
- Modify the current sample graph or load other graphs from your computer.
-
Right click a node and select "Set as root node" to start the decision tree at this
node. Right click the empty canvas to clear the root node.
-
Click the start button
to start the decision tree with the current graph.
Source Graph Structure
The DecisionTree component supports different formats of graphs.
-
The source graph does not have to be a tree. It can be an arbitrary graph structure.
- Loops in the graph are supported (see quiz sample).
-
The graph can contain group nodes, but the clickable nodes have to be normal nodes.
-
Only group nodes that are direct parents of clickable nodes appear in the decision
graph.
-
Edges connecting nodes in groups can either end directly at the child nodes (as in the
quiz sample), or at the containing group node (as in the what-to-do
sample).
The first node that's found that has no incoming edges serves as the start node. The
DecisionTree component also supports explicitly specifying a start node.