aalpy.utils.AutomatonGenerators
View Source
import random from aalpy.automata import Dfa, DfaState, MdpState, Mdp, MealyMachine, MealyState, \ MooreMachine, MooreState, OnfsmState, Onfsm, MarkovChain, McState, StochasticMealyState, StochasticMealyMachine from aalpy.utils.HelperFunctions import random_string_generator def generate_random_mealy_machine(num_states, input_alphabet, output_alphabet, compute_prefixes=False) -> MealyMachine: """ Generates a random Mealy machine. Args: num_states: number of states input_alphabet: input alphabet output_alphabet: output alphabet compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False) Returns: Mealy machine with num_states states """ states = list() for i in range(num_states): states.append(MealyState(i)) state_buffer = list(states) for state in states: for a in input_alphabet: if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[a] = new_state state.output_fun[a] = random.choice(output_alphabet) mm = MealyMachine(states[0], states) if compute_prefixes: for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm def generate_random_moore_machine(num_states, input_alphabet, output_alphabet, compute_prefixes=False) -> MooreMachine: """ Generates a random Moore machine. Args: num_states: number of states input_alphabet: input alphabet output_alphabet: output alphabet compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False) Returns: Moore machine with num_states states """ states = list() for i in range(num_states): states.append(MooreState(i, random.choice(output_alphabet))) state_buffer = list(states) for state in states: for a in input_alphabet: if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[a] = new_state mm = MooreMachine(states[0], states) if compute_prefixes: for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm def generate_random_dfa(num_states, alphabet, num_accepting_states=1, compute_prefixes=False) -> Dfa: """ Generates a random DFA. Args: num_states: number of states alphabet: input alphabet num_accepting_states: number of accepting states (Default value = 1) compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False) Returns: DFA """ assert num_states >= num_accepting_states states = list() for i in range(num_states): states.append(DfaState(i)) state_buffer = list(states) for state in states: for a in alphabet: if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[a] = new_state for _ in range(num_accepting_states): random.choice(states).is_accepting = True dfa = Dfa(states[0], states) if compute_prefixes: for state in states: state.prefix = dfa.get_shortest_path(dfa.initial_state, state) return dfa def generate_random_mdp(num_states, input_size, output_size, possible_probabilities=None): """ Generates random MDP. Args: num_states: number of states input_size: number of inputs output_size: user predefined outputs possible_probabilities: list of possible probability pairs to choose from Returns: random MDP """ inputs = [f'i{i+1}' for i in range(input_size)] outputs = [f'o{i+1}' for i in range(output_size)] if not possible_probabilities: possible_probabilities = [(1.,), (1.,), (1.,), (0.9, 0.1), (0.8, 0.2), (0.7, 0.3), (0.8, 0.1, 0.1), (0.7, 0.2, 0.1), (0.6, 0.2, 0.1, 0.1)] # ensure that there are no infinite loops possible_probabilities = [p for p in possible_probabilities if len(p) <= num_states] state_outputs = outputs.copy() states = [] for i in range(num_states): curr_output = state_outputs.pop(0) if state_outputs else random.choice(outputs) states.append(MdpState(f'q{i}', curr_output)) state_buffer = list(states) for state in states: for i in inputs: prob = random.choice(possible_probabilities) reached_states = [] for _ in prob: while True: new_state = random.choice(state_buffer) if state_buffer else random.choice(states) # ensure determinism if new_state.output not in {s.output for s in reached_states}: break if state_buffer: state_buffer.remove(new_state) reached_states.append(new_state) for prob, reached_state in zip(prob, reached_states): state.transitions[i].append((reached_state, prob)) return Mdp(states[0], states) def generate_random_smm(num_states, input_size, output_size, possible_probabilities=None): """ Generates random MDP. Args: num_states: number of states input_size: number of inputs output_size: number of outputs possible_probabilities: list of possible probability pairs to choose from Returns: random SMM """ inputs = [f'i{i + 1}' for i in range(input_size)] outputs = [f'o{i + 1}' for i in range(output_size)] if not possible_probabilities: possible_probabilities = [(1.,), (1.,), (1.,), (0.9, 0.1), (0.8, 0.2), (0.7, 0.3), (0.8, 0.1, 0.1), (0.7, 0.2, 0.1), (0.6, 0.2, 0.1, 0.1)] # ensure that there are no infinite loops possible_probabilities = [p for p in possible_probabilities if len(p) <= num_states] states = [] for i in range(num_states): states.append(StochasticMealyState(f'q{i}')) state_buffer = list(states) output_buffer = outputs.copy() for state in states: for i in inputs: prob = random.choice(possible_probabilities) reached_states = [] transition_outputs = [] for _ in prob: while True: o = random.choice(output_buffer) if output_buffer else random.choice(outputs) new_state = random.choice(state_buffer) if state_buffer else random.choice(states) # ensure determinism if o not in transition_outputs: break if output_buffer: output_buffer.remove(o) if state_buffer: state_buffer.remove(new_state) reached_states.append(new_state) transition_outputs.append(o) for index in range(len(prob)): state.transitions[i].append((reached_states[index], transition_outputs[index], prob[index])) return StochasticMealyMachine(states[0], states) def generate_random_ONFSM(num_states, num_inputs, num_outputs, multiple_out_prob=0.1): """ Randomly generate an observable non-deterministic finite-state machine. Args: num_states: number of states num_inputs: number of inputs num_outputs: number of outputs multiple_out_prob: probability that state will have multiple outputs (Default value = 0.5) Returns: randomly generated ONFSM """ inputs = [f'i{i+1}' for i in range(num_inputs)] outputs = [f'o{i+1}' for i in range(num_outputs)] states = [] for i in range(num_states): state = OnfsmState(f's{i}') states.append(state) state_buffer = states.copy() for state in states: for i in inputs: state_outputs = 1 if random.random() <= multiple_out_prob and num_outputs >= 2: state_outputs = random.randint(2, num_outputs) random_out = random.sample(outputs, state_outputs) for index in range(state_outputs): if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[i].append((random_out[index], new_state)) return Onfsm(states[0], states) def generate_random_markov_chain(num_states): assert num_states >= 3 possible_probabilities = [1.0, 1.0, 0.8, 0.5, 0.9] states = [] for i in range(num_states): states.append(McState(f'q{i}', i)) for index, state in enumerate(states[:-1]): prob = random.choice(possible_probabilities) if prob == 1.: new_state = states[index + 1] state.transitions.append((new_state, prob)) else: next_state = states[index + 1] up_states = list(states) up_states.remove(next_state) rand_state = random.choice(up_states) state.transitions.append((next_state, prob)) state.transitions.append((rand_state, round(1 - prob, 2))) return MarkovChain(states[0], states) def dfa_from_state_setup(state_setup) -> Dfa: """ First state in the state setup is the initial state. Example state setup: state_setup = { "a": (True, {"x": "b1", "y": "a"}), "b1": (False, {"x": "b2", "y": "a"}), "b2": (True, {"x": "b3", "y": "a"}), "b3": (False, {"x": "b4", "y": "a"}), "b4": (False, {"x": "c", "y": "a"}), "c": (True, {"x": "a", "y": "a"}), } Args: state_setup: map from state_id to tuple(output and transitions_dict) Returns: DFA """ # state_setup should map from state_id to tuple(is_accepting and transitions_dict) # build states with state_id and output states = {key: DfaState(key, val[0]) for key, val in state_setup.items()} # add transitions to states for state_id, state in states.items(): for _input, target_state_id in state_setup[state_id][1].items(): state.transitions[_input] = states[target_state_id] # states to list states = [state for state in states.values()] # build moore machine with first state as starting state dfa = Dfa(states[0], states) for state in states: state.prefix = dfa.get_shortest_path(dfa.initial_state, state) return dfa def moore_from_state_setup(state_setup) -> MooreMachine: """ First state in the state setup is the initial state. Example state setup: state_setup = { "a": ("a", {"x": "b1", "y": "a"}), "b1": ("b", {"x": "b2", "y": "a"}), "b2": ("b", {"x": "b3", "y": "a"}), "b3": ("b", {"x": "b4", "y": "a"}), "b4": ("b", {"x": "c", "y": "a"}), "c": ("c", {"x": "a", "y": "a"}), } Args: state_setup: map from state_id to tuple(output and transitions_dict) Returns: Moore machine """ # build states with state_id and output states = {key: MooreState(key, val[0]) for key, val in state_setup.items()} # add transitions to states for state_id, state in states.items(): for _input, target_state_id in state_setup[state_id][1].items(): state.transitions[_input] = states[target_state_id] # states to list states = [state for state in states.values()] # build moore machine with first state as starting state mm = MooreMachine(states[0], states) for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm def mealy_from_state_setup(state_setup) -> MealyMachine: """ First state in the state setup is the initial state. state_setup = { "a": {"x": ("o1", "b1"), "y": ("o2", "a")}, "b1": {"x": ("o3", "b2"), "y": ("o1", "a")}, "b2": {"x": ("o1", "b3"), "y": ("o2", "a")}, "b3": {"x": ("o3", "b4"), "y": ("o1", "a")}, "b4": {"x": ("o1", "c"), "y": ("o4", "a")}, "c": {"x": ("o3", "a"), "y": ("o5", "a")}, } Args: state_setup: state_setup should map from state_id to tuple(transitions_dict). Returns: Mealy Machine """ # state_setup should map from state_id to tuple(transitions_dict). # Each entry in transition dict is <input> : <output, new_state_id> # build states with state_id and output states = {key: MealyState(key) for key, _ in state_setup.items()} # add transitions to states for state_id, state in states.items(): for _input, (output, new_state) in state_setup[state_id].items(): state.transitions[_input] = states[new_state] state.output_fun[_input] = output # states to list states = [state for state in states.values()] # build moore machine with first state as starting state mm = MealyMachine(states[0], states) for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm
View Source
def generate_random_mealy_machine(num_states, input_alphabet, output_alphabet, compute_prefixes=False) -> MealyMachine: """ Generates a random Mealy machine. Args: num_states: number of states input_alphabet: input alphabet output_alphabet: output alphabet compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False) Returns: Mealy machine with num_states states """ states = list() for i in range(num_states): states.append(MealyState(i)) state_buffer = list(states) for state in states: for a in input_alphabet: if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[a] = new_state state.output_fun[a] = random.choice(output_alphabet) mm = MealyMachine(states[0], states) if compute_prefixes: for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm
Generates a random Mealy machine.
Args:
num_states: number of states
input_alphabet: input alphabet
output_alphabet: output alphabet
compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False)
Returns: Mealy machine with num_states states
View Source
def generate_random_moore_machine(num_states, input_alphabet, output_alphabet, compute_prefixes=False) -> MooreMachine: """ Generates a random Moore machine. Args: num_states: number of states input_alphabet: input alphabet output_alphabet: output alphabet compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False) Returns: Moore machine with num_states states """ states = list() for i in range(num_states): states.append(MooreState(i, random.choice(output_alphabet))) state_buffer = list(states) for state in states: for a in input_alphabet: if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[a] = new_state mm = MooreMachine(states[0], states) if compute_prefixes: for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm
Generates a random Moore machine.
Args:
num_states: number of states
input_alphabet: input alphabet
output_alphabet: output alphabet
compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False)
Returns:
Moore machine with num_states states
View Source
def generate_random_dfa(num_states, alphabet, num_accepting_states=1, compute_prefixes=False) -> Dfa: """ Generates a random DFA. Args: num_states: number of states alphabet: input alphabet num_accepting_states: number of accepting states (Default value = 1) compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False) Returns: DFA """ assert num_states >= num_accepting_states states = list() for i in range(num_states): states.append(DfaState(i)) state_buffer = list(states) for state in states: for a in alphabet: if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[a] = new_state for _ in range(num_accepting_states): random.choice(states).is_accepting = True dfa = Dfa(states[0], states) if compute_prefixes: for state in states: state.prefix = dfa.get_shortest_path(dfa.initial_state, state) return dfa
Generates a random DFA.
Args:
num_states: number of states
alphabet: input alphabet
num_accepting_states: number of accepting states (Default value = 1)
compute_prefixes: if true, shortest path to reach each state will be computed (Default value = False)
Returns:
DFA
View Source
def generate_random_mdp(num_states, input_size, output_size, possible_probabilities=None): """ Generates random MDP. Args: num_states: number of states input_size: number of inputs output_size: user predefined outputs possible_probabilities: list of possible probability pairs to choose from Returns: random MDP """ inputs = [f'i{i+1}' for i in range(input_size)] outputs = [f'o{i+1}' for i in range(output_size)] if not possible_probabilities: possible_probabilities = [(1.,), (1.,), (1.,), (0.9, 0.1), (0.8, 0.2), (0.7, 0.3), (0.8, 0.1, 0.1), (0.7, 0.2, 0.1), (0.6, 0.2, 0.1, 0.1)] # ensure that there are no infinite loops possible_probabilities = [p for p in possible_probabilities if len(p) <= num_states] state_outputs = outputs.copy() states = [] for i in range(num_states): curr_output = state_outputs.pop(0) if state_outputs else random.choice(outputs) states.append(MdpState(f'q{i}', curr_output)) state_buffer = list(states) for state in states: for i in inputs: prob = random.choice(possible_probabilities) reached_states = [] for _ in prob: while True: new_state = random.choice(state_buffer) if state_buffer else random.choice(states) # ensure determinism if new_state.output not in {s.output for s in reached_states}: break if state_buffer: state_buffer.remove(new_state) reached_states.append(new_state) for prob, reached_state in zip(prob, reached_states): state.transitions[i].append((reached_state, prob)) return Mdp(states[0], states)
Generates random MDP.
Args:
num_states: number of states
input_size: number of inputs
output_size: user predefined outputs
possible_probabilities: list of possible probability pairs to choose from
Returns:
random MDP
View Source
def generate_random_smm(num_states, input_size, output_size, possible_probabilities=None): """ Generates random MDP. Args: num_states: number of states input_size: number of inputs output_size: number of outputs possible_probabilities: list of possible probability pairs to choose from Returns: random SMM """ inputs = [f'i{i + 1}' for i in range(input_size)] outputs = [f'o{i + 1}' for i in range(output_size)] if not possible_probabilities: possible_probabilities = [(1.,), (1.,), (1.,), (0.9, 0.1), (0.8, 0.2), (0.7, 0.3), (0.8, 0.1, 0.1), (0.7, 0.2, 0.1), (0.6, 0.2, 0.1, 0.1)] # ensure that there are no infinite loops possible_probabilities = [p for p in possible_probabilities if len(p) <= num_states] states = [] for i in range(num_states): states.append(StochasticMealyState(f'q{i}')) state_buffer = list(states) output_buffer = outputs.copy() for state in states: for i in inputs: prob = random.choice(possible_probabilities) reached_states = [] transition_outputs = [] for _ in prob: while True: o = random.choice(output_buffer) if output_buffer else random.choice(outputs) new_state = random.choice(state_buffer) if state_buffer else random.choice(states) # ensure determinism if o not in transition_outputs: break if output_buffer: output_buffer.remove(o) if state_buffer: state_buffer.remove(new_state) reached_states.append(new_state) transition_outputs.append(o) for index in range(len(prob)): state.transitions[i].append((reached_states[index], transition_outputs[index], prob[index])) return StochasticMealyMachine(states[0], states)
Generates random MDP.
Args:
num_states: number of states
input_size: number of inputs
output_size: number of outputs
possible_probabilities: list of possible probability pairs to choose from
Returns:
random SMM
View Source
def generate_random_ONFSM(num_states, num_inputs, num_outputs, multiple_out_prob=0.1): """ Randomly generate an observable non-deterministic finite-state machine. Args: num_states: number of states num_inputs: number of inputs num_outputs: number of outputs multiple_out_prob: probability that state will have multiple outputs (Default value = 0.5) Returns: randomly generated ONFSM """ inputs = [f'i{i+1}' for i in range(num_inputs)] outputs = [f'o{i+1}' for i in range(num_outputs)] states = [] for i in range(num_states): state = OnfsmState(f's{i}') states.append(state) state_buffer = states.copy() for state in states: for i in inputs: state_outputs = 1 if random.random() <= multiple_out_prob and num_outputs >= 2: state_outputs = random.randint(2, num_outputs) random_out = random.sample(outputs, state_outputs) for index in range(state_outputs): if state_buffer: new_state = random.choice(state_buffer) state_buffer.remove(new_state) else: new_state = random.choice(states) state.transitions[i].append((random_out[index], new_state)) return Onfsm(states[0], states)
Randomly generate an observable non-deterministic finite-state machine.
Args:
num_states: number of states num_inputs: number of inputs num_outputs: number of outputs multiple_out_prob: probability that state will have multiple outputs (Default value = 0.5)
Returns:
randomly generated ONFSM
View Source
def generate_random_markov_chain(num_states): assert num_states >= 3 possible_probabilities = [1.0, 1.0, 0.8, 0.5, 0.9] states = [] for i in range(num_states): states.append(McState(f'q{i}', i)) for index, state in enumerate(states[:-1]): prob = random.choice(possible_probabilities) if prob == 1.: new_state = states[index + 1] state.transitions.append((new_state, prob)) else: next_state = states[index + 1] up_states = list(states) up_states.remove(next_state) rand_state = random.choice(up_states) state.transitions.append((next_state, prob)) state.transitions.append((rand_state, round(1 - prob, 2))) return MarkovChain(states[0], states)
View Source
def dfa_from_state_setup(state_setup) -> Dfa: """ First state in the state setup is the initial state. Example state setup: state_setup = { "a": (True, {"x": "b1", "y": "a"}), "b1": (False, {"x": "b2", "y": "a"}), "b2": (True, {"x": "b3", "y": "a"}), "b3": (False, {"x": "b4", "y": "a"}), "b4": (False, {"x": "c", "y": "a"}), "c": (True, {"x": "a", "y": "a"}), } Args: state_setup: map from state_id to tuple(output and transitions_dict) Returns: DFA """ # state_setup should map from state_id to tuple(is_accepting and transitions_dict) # build states with state_id and output states = {key: DfaState(key, val[0]) for key, val in state_setup.items()} # add transitions to states for state_id, state in states.items(): for _input, target_state_id in state_setup[state_id][1].items(): state.transitions[_input] = states[target_state_id] # states to list states = [state for state in states.values()] # build moore machine with first state as starting state dfa = Dfa(states[0], states) for state in states: state.prefix = dfa.get_shortest_path(dfa.initial_state, state) return dfa
First state in the state setup is the initial state. Example state setup: state_setup = { "a": (True, {"x": "b1", "y": "a"}), "b1": (False, {"x": "b2", "y": "a"}), "b2": (True, {"x": "b3", "y": "a"}), "b3": (False, {"x": "b4", "y": "a"}), "b4": (False, {"x": "c", "y": "a"}), "c": (True, {"x": "a", "y": "a"}), }
Args:
state_setup: map from state_id to tuple(output and transitions_dict)
Returns:
DFA
View Source
def moore_from_state_setup(state_setup) -> MooreMachine: """ First state in the state setup is the initial state. Example state setup: state_setup = { "a": ("a", {"x": "b1", "y": "a"}), "b1": ("b", {"x": "b2", "y": "a"}), "b2": ("b", {"x": "b3", "y": "a"}), "b3": ("b", {"x": "b4", "y": "a"}), "b4": ("b", {"x": "c", "y": "a"}), "c": ("c", {"x": "a", "y": "a"}), } Args: state_setup: map from state_id to tuple(output and transitions_dict) Returns: Moore machine """ # build states with state_id and output states = {key: MooreState(key, val[0]) for key, val in state_setup.items()} # add transitions to states for state_id, state in states.items(): for _input, target_state_id in state_setup[state_id][1].items(): state.transitions[_input] = states[target_state_id] # states to list states = [state for state in states.values()] # build moore machine with first state as starting state mm = MooreMachine(states[0], states) for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm
First state in the state setup is the initial state. Example state setup: state_setup = { "a": ("a", {"x": "b1", "y": "a"}), "b1": ("b", {"x": "b2", "y": "a"}), "b2": ("b", {"x": "b3", "y": "a"}), "b3": ("b", {"x": "b4", "y": "a"}), "b4": ("b", {"x": "c", "y": "a"}), "c": ("c", {"x": "a", "y": "a"}), }
Args:
state_setup: map from state_id to tuple(output and transitions_dict)
Returns:
Moore machine
View Source
def mealy_from_state_setup(state_setup) -> MealyMachine: """ First state in the state setup is the initial state. state_setup = { "a": {"x": ("o1", "b1"), "y": ("o2", "a")}, "b1": {"x": ("o3", "b2"), "y": ("o1", "a")}, "b2": {"x": ("o1", "b3"), "y": ("o2", "a")}, "b3": {"x": ("o3", "b4"), "y": ("o1", "a")}, "b4": {"x": ("o1", "c"), "y": ("o4", "a")}, "c": {"x": ("o3", "a"), "y": ("o5", "a")}, } Args: state_setup: state_setup should map from state_id to tuple(transitions_dict). Returns: Mealy Machine """ # state_setup should map from state_id to tuple(transitions_dict). # Each entry in transition dict is <input> : <output, new_state_id> # build states with state_id and output states = {key: MealyState(key) for key, _ in state_setup.items()} # add transitions to states for state_id, state in states.items(): for _input, (output, new_state) in state_setup[state_id].items(): state.transitions[_input] = states[new_state] state.output_fun[_input] = output # states to list states = [state for state in states.values()] # build moore machine with first state as starting state mm = MealyMachine(states[0], states) for state in states: state.prefix = mm.get_shortest_path(mm.initial_state, state) return mm
First state in the state setup is the initial state. state_setup = { "a": {"x": ("o1", "b1"), "y": ("o2", "a")}, "b1": {"x": ("o3", "b2"), "y": ("o1", "a")}, "b2": {"x": ("o1", "b3"), "y": ("o2", "a")}, "b3": {"x": ("o3", "b4"), "y": ("o1", "a")}, "b4": {"x": ("o1", "c"), "y": ("o4", "a")}, "c": {"x": ("o3", "a"), "y": ("o5", "a")}, }
Args:
state_setup:
state_setup should map from state_id to tuple(transitions_dict).
Returns:
Mealy Machine