State Transition Table

[laniakea]

Is it possible to fetch a state transition table in a matrix format? I was unable to find anything regarding this.

[aleneum]

Hi @laniakea,

Option A: Iterate over machine.events

transitions sorts Events by their 'trigger'.
Each Event has a list of Transitions that are sorted by source.
Each Transition has a dest attribute that contains the name of the destination state.
(Optional: A Transition also contains the conditions that must be met for a it to happen. Just in case your use case is more complex)

If you iterate over events and transition you could collect a dictionary of trigger dictionaries.
This data structure could be used to create a representation that suits your needs.
I used pandas in the example below out of lazyness:

from transitions import Machine
from collections import defaultdict
from pprint import pprint
from pandas import DataFrame

states = ["A", "B", "C"]
transitions = [["go", "A", "B"], ["go", "B", "C"], ["reset", "*", "A"]]
data = defaultdict(lambda: defaultdict(list))
machine = Machine(states=states, transitions=transitions, initial="A")

for trigger, event in machine.events.items():
    for source, trans in event.transitions.items():
        for tran in trans:
            data[source][tran.dest].append(trigger)

pprint(data)  # [1]

print(DataFrame([[", ".join(data[source][dest]) for dest in states] for source in states],
                columns=states, index=states))  # [2]

pprint [1]

defaultdict(<function <lambda> at 0x100b923e0>,
            {'A': defaultdict(<class 'list'>,
                              {'A': ['to_A', 'reset'],
                               'B': ['to_B', 'go'],
                               'C': ['to_C']}),
             'B': defaultdict(<class 'list'>,
                              {'A': ['to_A', 'reset'],
                               'B': ['to_B'],
                               'C': ['to_C', 'go']}),
             'C': defaultdict(<class 'list'>,
                              {'A': ['to_A', 'reset'],
                               'B': ['to_B'],
                               'C': ['to_C']})})

Pandas [2]

             A         B         C
A  to_A, reset  to_B, go      to_C
B  to_A, reset      to_B  to_C, go
C  to_A, reset      to_B      to_C

Option B: Iterate over pygraphviz.AGraph

Alternatively, you could use a graph created by the diagram extension. If you use the pygraphviz backend, you can iterate over all edges and nodes and get a similar result but with conditions:

from transitions.extensions import GraphMachine
from collections import defaultdict
from pygraphviz import AGraph
from pandas import DataFrame

states = ["A", "B", "C"]
transitions = [["go", "A", "B"], ["go", "B", "C"], ["reset", "*", "A"]]
data = defaultdict(lambda: defaultdict(lambda: "-"))
machine = GraphMachine(states=states, transitions=transitions, initial="A", show_auto_transitions=False,
                       show_conditions=True, graph_engine="pygraphviz")
machine.add_transition("go", "C", "B", conditions="is_it_friday")
machine.add_transition("go", "C", "A")
machine.add_transition("stop", "C", "C", unless="not_done")

graph: AGraph = machine.get_graph()
for source in states:
    for edge in graph.edges_iter(source):
        if edge[0] == source:
            data[source][edge[1]] = edge.attr["label"]

print(DataFrame([[data[source][dest] for dest in states] for source in states],
                columns=states, index=states))  # [3]

Pandas [3]

            A                  B                 C
A       reset                 go                 -
B       reset                  -                go
C  go | reset  go [is_it_friday]  stop [!not_done]

[abdolreza-sh]

Many thanks for the comprehensive answer