gillespy2/solvers/numpy/tau_leaping_solver.py
# GillesPy2 is a modeling toolkit for biochemical simulation.
# Copyright (C) 2019-2023 GillesPy2 developers.
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
"""Class and methods for the Tau Leaping Solver"""
import copy
import random
import math
from threading import Thread, Event
import numpy as np
from gillespy2.solvers.utilities import Tau
from gillespy2.solvers.utilities import solverutils as nputils
from gillespy2.core import GillesPySolver, log, liveGraphing
from gillespy2.core.gillespyError import *
from gillespy2.core.results import Results
class TauLeapingSolver(GillesPySolver):
"""
A Tau Leaping solver for GillesPy2 models. This solver uses an algorithm that calculates
multiple reactions in a single step over a given tau step size. The change in propensities
over this step are bounded by bounding the relative change in state, yielding greatly improved
run-time performance with very little trade-off in accuracy.
:param model: The model on which the solver will operate.
:type model: gillespy2.Model
"""
name = "TauLeapingSolver"
rc = 0
stop_event = None
pause_event = None
result = None
def __init__(self, model=None, debug=False, constant_tau_stepsize=None):
if model is None:
raise SimulationError("A model is required to run the simulation.")
name = "TauLeapingSolver"
rc = 0
stop_event = None
pause_event = None
result = None
self.model = copy.deepcopy(model)
self.debug = debug
self.is_instantiated = True
self.constant_tau_stepsize = constant_tau_stepsize
def __get_reactions(self, step, curr_state, curr_time, save_time, propensities, reactions):
"""
Helper Function to get reactions fired from t to t+tau.
:returns: Three values:
rxn_count - dict with key=Reaction channel value=number of times fired
curr_state - dict containing all state variables for system at current time
curr_time - float representing current time
"""
if curr_time + step > save_time:
if self.debug:
print("Step exceeds save_time, changing step size from ", step,
" to ", save_time - curr_time)
step = save_time - curr_time
if self.debug:
print("Curr Time: ", curr_time, " Save time: ", save_time, "step: ", step)
rxn_count = {}
for rxn in reactions:
rxn_count[rxn] = np.random.poisson(propensities[rxn] * step)
if self.debug:
print("Reactions Fired: ", rxn_count)
curr_time = curr_time+step
return rxn_count, curr_state, curr_time
@classmethod
def get_solver_settings(cls):
"""
Returns a list of arguments supported by tau_leaping_solver.run.
:returns: Tuple of strings, denoting all keyword argument for this solvers run() method.
:rtype: tuple
"""
return ('model', 't', 'number_of_trajectories', 'increment', 'seed', 'debug', 'profile','timeout', 'tau_tol')
def run(self=None, model=None, t=None, number_of_trajectories=1, increment=None, seed=None,
debug=False, profile=False, live_output=None, live_output_options={},
timeout=None, resume=None, tau_tol=0.03, **kwargs):
"""
Function calling simulation of the model.
This is typically called by the run function in GillesPy2 model objects
and will inherit those parameters which are passed with the model
as the arguments this run function.
:param model: The model on which the solver will operate. (Deprecated)
:type model: gillespy2.Model
:param t: Simulation run time.
:type t: int or float
:param number_of_trajectories: Number of trajectories to simulate. By default number_of_trajectories = 1.
:type number_of_trajectories: int
:param increment: Save point increment for recording data.
:type increment: float
:param seed: The random seed for the simulation. Optional, defaults to None.
:type seed: int
:param debug: Set to True to provide additional debug information about the simulation.
:type debug: bool
:param profile: Set to True to provide information about step size (tau) taken at each step.
:type profile: bool
:param live_output: The type of output to be displayed by solver. Can be "progress", "text", or "graph".
:type live_output: str
:param live_output_options: Contains options for live_output. By default {"interval":1}. "interval"
specifies seconds between displaying. "clear_output" specifies if display should be refreshed with each
display.
:type live_output_options: dict
:param timeout: If set, if simulation takes longer than timeout, will exit.
:type timeout: int
:param resume: Result of a previously run simulation, to be resumed.
:type resume: gillespy2.Results
:param tau_tol: Tolerance level for Tau leaping algorithm. Larger tolerance values will
result in larger tau steps. Default value is 0.03.
:type tau_tol: float
:returns: A result object containing the results of the simulation.
:rtype: gillespy2.Results
"""
from gillespy2 import log
if self is None:
# Post deprecation block
# raise SimulationError("TauLeapingSolver must be instantiated to run the simulation")
# Pre deprecation block
log.warning(
"""
`gillespy2.Model.run(solver=TauLeapingSolver)` is deprecated.
You should use `gillespy2.Model.run(solver=TauLeapingSolver(model=gillespy2.Model))
Future releases of GillesPy2 may not support this feature.
"""
)
self = TauLeapingSolver(model=model, debug=debug, profile=profile)
if model is not None:
log.warning('model = gillespy2.model is deprecated. Future releases '
'of GillesPy2 may not support this feature.')
if self.model is None:
if model is None:
raise SimulationError("A model is required to run the simulation.")
self.model = copy.deepcopy(model)
self.model.compile_prep()
self.validate_model(self.model, model)
self.validate_sbml_features(model=self.model)
self.validate_tspan(increment=increment, t=t)
if increment is None:
increment = self.model.tspan[-1] - self.model.tspan[-2]
if t is None:
t = self.model.tspan[-1]
self.stop_event = Event()
self.pause_event = Event()
if timeout is not None and timeout <= 0:
timeout = None
if len(kwargs) > 0:
for key in kwargs:
log.warning('Unsupported keyword argument to {0} solver: {1}'.format(self.name, key))
# create numpy array for timeline
if resume is not None:
# start where we last left off if resuming a simulatio
lastT = resume['time'][-1]
step = lastT - resume['time'][-2]
timeline = np.arange(lastT, t+step, step)
else:
timeline = np.linspace(0, t, int(round(t / increment + 1)))
species = list(self.model._listOfSpecies.keys())
trajectory_base, tmpSpecies = nputils.numpy_trajectory_base_initialization(self.model, number_of_trajectories,
timeline, species, resume=resume)
# total_time and curr_state are list of len 1 so that __run receives reference
if resume is not None:
total_time = [resume['time'][-1]]
else:
total_time = [0]
curr_state = [None]
live_grapher = [None]
sim_thread = Thread(target=self.___run, args=(curr_state, total_time, timeline, trajectory_base, tmpSpecies,
live_grapher,), kwargs={'t': t,
'number_of_trajectories':
number_of_trajectories,
'increment': increment, 'seed': seed,
'debug': debug, 'resume': resume,
'timeout': timeout, 'tau_tol': tau_tol
})
try:
time = 0
sim_thread.start()
if live_output is not None:
import gillespy2.core.liveGraphing
live_output_options['type'] = live_output
gillespy2.core.liveGraphing.valid_graph_params(
live_output_options)
if resume is not None:
resumeTest = True # If resuming, relay this information to live_grapher
else:
resumeTest = False
live_grapher[
0] = gillespy2.core.liveGraphing.LiveDisplayer(self.model,
timeline,
number_of_trajectories,
live_output_options,
resume=resumeTest)
display_timer = gillespy2.core.liveGraphing.RepeatTimer(
live_output_options['interval'],
live_grapher[0].display, args=(curr_state,
total_time,
trajectory_base,
live_output
)
)
display_timer.start()
if timeout is not None:
while sim_thread.is_alive():
sim_thread.join(.1)
time += .1
if time >= timeout:
break
else:
while sim_thread.is_alive():
sim_thread.join(.1)
if live_grapher[0] is not None:
display_timer.cancel()
self.stop_event.set()
while self.result is None:
pass
except KeyboardInterrupt:
if live_output:
display_timer.pause = True
display_timer.cancel()
self.pause_event.set()
while self.result is None:
pass
if hasattr(self, 'has_raised_exception'):
raise SimulationError(
f"Error encountered while running simulation:\nReturn code: {int(self.rc)}.\n"
) from self.has_raised_exception
return Results.build_from_solver_results(self, live_output_options)
def ___run(self, curr_state,total_time, timeline, trajectory_base, tmpSpecies, live_grapher, t=20,
number_of_trajectories=1, increment=0.05, seed=None, debug=False, profile=False,
resume=None, tau_tol=0.03, **kwargs):
try:
self.__run(curr_state, total_time, timeline, trajectory_base, tmpSpecies, live_grapher, t, number_of_trajectories,
increment, seed, debug, profile, resume, tau_tol, **kwargs)
except Exception as e:
self.has_raised_exception = e
self.result = []
return [], -1
def __run(self, curr_state, total_time, timeline, trajectory_base, tmpSpecies, live_grapher, t=20,
number_of_trajectories=1, increment=0.05, seed=None, debug=False, profile=False,
resume=None, tau_tol=0.03, **kwargs):
# for use with resume, determines how much excess data to cut off due to
# how species and time are initialized to 0
timeStopped = 0
if resume is not None:
if resume[0].model != self.model:
raise ModelError('When resuming, one must not alter the model being resumed.')
if t < resume['time'][-1]:
raise ExecutionError(
"'t' must be greater than previous simulations end time, or set in the run() method as the "
"simulations next end time")
if debug:
print("t = ", t)
print("increment = ", increment)
species_mappings, species, parameter_mappings, number_species = nputils.numpy_initialization(self.model)
if seed is not None:
if not isinstance(seed, int):
seed = int(seed)
if seed > 0:
random.seed(seed)
np.random.seed(seed)
else:
raise ModelError('seed must be a positive integer')
simulation_data = []
for trajectory_num in range(number_of_trajectories):
if self.stop_event.is_set():
self.rc = 33
break
elif self.pause_event.is_set():
timeStopped = timeline[entry_count]
# For multi trajectories, live_grapher needs to be informed of trajectory increment
if live_grapher[0] is not None:
live_grapher[0].increment_trajectory(trajectory_num)
start_state = [0] * (len(self.model.listOfReactions) + len(self.model.listOfRateRules))
propensities = {}
curr_state[0] = {}
if resume is not None:
save_time = total_time[0]
curr_time = [save_time]
else:
save_time = 0
curr_time = [0]
curr_state[0]['vol'] = self.model.volume
data = {'time': timeline}
steps_taken = []
steps_rejected = 0
entry_count = 0
trajectory = trajectory_base[trajectory_num]
HOR, reactants, mu_i, sigma_i, g_i, epsilon_i, critical_threshold = Tau.initialize(self.model, tau_tol)
if resume is not None:
for spec in self.model.listOfSpecies:
curr_state[0][spec] = tmpSpecies[spec]
else:
for spec in self.model.listOfSpecies:
curr_state[0][spec] = self.model.listOfSpecies[spec].initial_value
for param in self.model.listOfParameters:
curr_state[0][param] = self.model.listOfParameters[param].value
for i, rxn in enumerate(self.model.listOfReactions):
# set reactions to uniform random number and add to start_state
start_state[i] = (math.log(random.uniform(0, 1)))
if debug:
print("Setting Random number ",
start_state[i], " for ", self.model.listOfReactions[rxn].name)
compiled_propensities = {}
for i, r in enumerate(self.model.listOfReactions):
compiled_propensities[r] = compile(self.model.listOfReactions[r].propensity_function, '<string>', 'eval')
timestep = 0
# Each save step
while entry_count < timeline.size:
if self.stop_event.is_set():
self.rc = 33
break
elif self.pause_event.is_set():
timeStopped = timeline[entry_count]
break
# Until save step reached
while curr_time[0] < save_time:
if self.stop_event.is_set():
self.rc = 33
break
elif self.pause_event.is_set():
timeStopped = timeline[entry_count]
break
propensity_sum = 0
for i, r in enumerate(self.model.listOfReactions):
propensities[r] = eval(compiled_propensities[r], curr_state[0])
propensity_sum += propensities[r]
tau_args = [HOR, reactants, mu_i, sigma_i, g_i, epsilon_i, tau_tol, critical_threshold,
self.model, propensities, curr_state[0], curr_time[0], save_time]
if self.constant_tau_stepsize is None:
tau_step = Tau.select(*tau_args)
#print(f"t={curr_time[0]} tau={tau_step}")
else:
tau_step = self.constant_tau_stepsize
prev_start_state = start_state.copy()
prev_curr_state = curr_state[0].copy()
prev_curr_time = curr_time[0]
total_time[0] = curr_time[0]
loop_cnt = 0
while True:
loop_cnt += 1
if loop_cnt > 100:
raise Exception("Loop over __get_reactions() exceeded loop count")
reactions, curr_state[0], curr_time[0] = self.__get_reactions(
tau_step, curr_state[0], curr_time[0], save_time,
propensities, self.model.listOfReactions)
# Update curr_state with the result of the SSA reaction that fired
species_modified = {}
for i, rxn in enumerate(self.model.listOfReactions):
if reactions[rxn] > 0:
for reactant in self.model.listOfReactions[rxn].reactants:
species_modified[reactant.name] = True
curr_state[0][reactant.name] -= self.model.listOfReactions[
rxn].reactants[reactant] * reactions[rxn]
for product in self.model.listOfReactions[rxn].products:
species_modified[product.name] = True
curr_state[0][product.name] += self.model.listOfReactions[
rxn].products[product] * reactions[rxn]
neg_state = False
for spec in species_modified:
if curr_state[0][spec] < 0:
neg_state = True
if debug:
print("Negative state detected: curr_state[{0}]= {1}".format(spec,
curr_state[0][spec]))
if neg_state:
if debug:
print("\trxn={0}".format(reactions))
start_state = prev_start_state.copy()
curr_state[0] = prev_curr_state.copy()
curr_time[0] = prev_curr_time
total_time[0] = prev_curr_time
tau_step = tau_step / 2
steps_rejected += 1
if debug:
print("Resetting curr_state[{0}]= {1}".format(spec, curr_state[0][spec]))
if debug:
print("\tRejecting step, taking step of half size, tau_step={0}".format(tau_step))
else:
break # breakout of the while True
# save step reached
for i in range(number_species):
trajectory[entry_count][i + 1] = curr_state[0][species[i]]
save_time += increment
timestep += 1
entry_count += 1
# end of trajectory
for i in range(number_species):
data[species[i]] = trajectory[:, i+1]
simulation_data.append(data)
if profile:
print(steps_taken)
print("Total Steps Taken: ", len(steps_taken))
print("Total Steps Rejected: ", steps_rejected)
# If simulation has been paused, or tstopped !=0
if timeStopped != 0 or resume is not None:
simulation_data = nputils.numpy_resume(timeStopped, simulation_data, resume=resume)
self.result = simulation_data
return self.result, self.rc