def multiprocessing(self):
print("\nDay = %s, trials = %s,..." %(self.day,self.n_trials))
pool = pathos.multiprocessing.ProcessingPool()
freeze_support()
exp_params = []
for drug in self.drugs:
for trial in self.trials:
exp_params.append([self.decision_type, self.drug_type, drug, trial, self.seed, self.P])
#self.df_list = []
self.df_list = pool.map(self.run, exp_params)
#for i in range(self.n_trials):self.df_list.append(self.run(exp_params[i]))
for i in range(len(self.df_list)):
self.accuracy_data.append(self.df_list[i][2])
self.accuracy_data = np.array(self.accuracy_data)
for i in range(len(self.df_list)):
self.spiking_array.append((self.df_list[i][3]))
self.spiking_array = np.array(self.spiking_array)
primary_dataframe = pd.concat([self.df_list[i][0] for i in range(len(self.df_list))], ignore_index=True)
firing_dataframe = pd.concat([self.df_list[i][1] for i in range(len(self.df_list))], ignore_index=True)
return primary_dataframe, firing_dataframe
def optimizeDE(cost,
xmin,
xmax,
npop=40,
maxiter=1000,
maxfun=1e+6,
convergence_tol=1e-5,
ngen=100,
crossover=0.9,
percent_change=0.9,
MINMAX=1,
x0=[],
radius=0.2,
parallel=False,
nodes=16):
from mystic.solvers import DifferentialEvolutionSolver2
from mystic.termination import ChangeOverGeneration as COG
from mystic.strategy import Best1Exp
from mystic.monitors import VerboseMonitor, Monitor
from pathos.helpers import freeze_support
freeze_support() # help Windows use multiprocessing
stepmon = VerboseMonitor(20) # step for showing live update
evalmon = Monitor()
ndim = len(xmin)
solver = DifferentialEvolutionSolver2(ndim, npop)
if parallel:
solver.SetMapper(Pool(nodes).map)
if x0 == []:
solver.SetRandomInitialPoints(xmin, xmax)
else:
solver.SetInitialPoints(x0, radius)
solver.SetStrictRanges(xmin, xmax)
solver.SetEvaluationLimits(maxiter, maxfun)
solver.SetEvaluationMonitor(evalmon)
solver.SetGenerationMonitor(stepmon)
tol = convergence_tol
solver.Solve(cost,
termination=COG(tolerance=tol, generations=ngen),
strategy=Best1Exp,
CrossProbability=crossover,
ScalingFactor=percent_change)
solved = solver.bestSolution
func_max = MINMAX * solver.bestEnergy
func_evals = solver.evaluations
return solved, func_max, func_evals
from mystic.math import poly1d
#from raw_chebyshev8 import chebyshev8cost as ChebyshevCost # no globals
#from raw_chebyshev8b import chebyshev8cost as ChebyshevCost # use globals
from mystic.models.poly import chebyshev8cost as ChebyshevCost # no helper
ND = 9
NP = 80
MAX_GENERATIONS = NP * NP
NNODES = NP // 5
seed = 321
if __name__ == '__main__':
from pathos.helpers import freeze_support
freeze_support() # help Windows use multiprocessing
def print_solution(func):
print(poly1d(func))
return
psow = VerboseMonitor(10)
ssow = VerboseMonitor(10)
random_seed(seed)
print("first sequential...")
solver = DifferentialEvolutionSolver2(ND, NP)
solver.SetRandomInitialPoints(min=[-100.0] * ND, max=[100.0] * ND)
solver.SetEvaluationLimits(generations=MAX_GENERATIONS)
solver.SetGenerationMonitor(ssow)
solver.Solve(ChebyshevCost, VTR(0.01), strategy=Best1Exp, \
pool._id = old_id
res = pool.map(squared, range(2))
assert res == [0, 1]
pool.terminate()
pool.clear()
assert len(state) == 0
return
def test_basic():
check_basic(ThreadPool(), tstate)
# check_basic(ProcessPool(), mstate)
# check_basic(ParallelPool(), pstate)
def test_rename():
check_rename(ThreadPool(), tstate)
check_rename(ProcessPool(), mstate)
check_rename(ParallelPool(), pstate)
def test_nodes():
check_nodes(ThreadPool(), tstate)
check_nodes(ProcessPool(), mstate)
check_nodes(ParallelPool(), pstate)
if __name__ == '__main__':
from pathos.helpers import freeze_support
freeze_support()
test_basic()
test_rename()
test_nodes()
assert res == std
def test_processing():
from pathos.pools import ProcessPool as MPP
pool = MPP()
res = timed_pool(pool, items, delay, verbose)
assert res == std
def test_threading():
from pathos.pools import ThreadPool as MTP
pool = MTP()
res = timed_pool(pool, items, delay, verbose)
assert res == std
if __name__ == '__main__':
if verbose:
print("CONFIG: delay = %s" % delay)
print("CONFIG: items = %s" % items)
print("")
from pathos.helpers import freeze_support, shutdown
freeze_support()
test_serial()
test_pp()
test_processing()
test_threading()
shutdown()
#from pathos.pools import ProcessPool as Pool
from pathos.pools import ParallelPool as Pool
solver = DifferentialEvolutionSolver2(ND, NP)
solver.SetMapper(Pool().map)
solver.SetRandomInitialPoints(min = [-100.0]*ND, max = [100.0]*ND)
solver.SetEvaluationLimits(generations=MAX_GENERATIONS)
solver.SetGenerationMonitor(VerboseMonitor(30))
solver.enable_signal_handler()
strategy = Best1Exp
#strategy = Best1Bin
solver.Solve(ChebyshevCost, termination=VTR(0.01), strategy=strategy, \
CrossProbability=1.0, ScalingFactor=0.9 , \
sigint_callback=plot_solution)
solution = solver.Solution()
return solution
if __name__ == '__main__':
from pathos.helpers import freeze_support
freeze_support() # help Windows use multiprocessing
solution = main()
print_solution(solution)
plot_solution(solution)
# end of file
'''
Command prompt yes or no question
'''
while "the answer is invalid":
reply = str(input(question + ' (y/n): ')).lower().strip()
if reply[:1] == 'y':
return True
if reply[:1] == 'n':
return False
print("invalid answer")
# end _yes_or_no
if __name__ == "__main__":
freeze_support(
) # To prevent pathos from having issues when run from windows command line
parser = argparse.ArgumentParser()
parser.add_argument("path", help="Path to the spectral data file")
parser.add_argument(
"-sf",
"--savefile",
help=
"Path to directory to save output files, by default will save to local calculations directory"
)
parser.add_argument(
"-w1",
"--width1",
type=float,
help="The beginning of the range of absorber widths (default 5 nm)")
parser.add_argument(
def main():
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import numpy as np
from helper import make_cues, make_addon, ch_dir, empirical_dataframe
from pathos.multiprocessing import ProcessingPool as Pool
from pathos.helpers import freeze_support #for Windows
# import ipdb
'''Import Parameters from File'''
P=eval(open('parameters.txt').read()) #parameter dictionary
seed=P['seed'] #sets tuning curves equal to control before drug application
n_trials=P['n_trials']
n_processes=P['n_processes']
drug_type=str(P['drug_type'])
decision_type=str(P['decision_type'])
drugs=P['drugs']
trials, perceived, cues = make_cues(P)
P['timesteps']=np.arange(0,int((P['t_cue']+P['t_delay'])/P['dt_sample']))
P['cues']=cues
P['perceived']=perceived
'''Multiprocessing'''
print "drug_type=%s, decision_type=%s, trials=%s..." %(drug_type,decision_type,n_trials)
freeze_support()
pool = Pool(nodes=n_processes)
exp_params=[]
for drug in drugs:
for trial in trials:
exp_params.append([decision_type, drug_type, drug, trial, seed, P])
# df_list=[run(exp_params[0]),run(exp_params[-1])] #for debugging, pool.map has bad traceback
df_list = pool.map(run, exp_params)
primary_dataframe = pd.concat([df_list[i][0] for i in range(len(df_list))], ignore_index=True)
firing_dataframe = pd.concat([df_list[i][1] for i in range(len(df_list))], ignore_index=True)
'''Plot and Export'''
print 'Exporting Data...'
datadir=ch_dir()
primary_dataframe.to_pickle('primary_data.pkl')
firing_dataframe.to_pickle('firing_data.pkl')
param_df=pd.DataFrame([P])
param_df.reset_index().to_json('params.json',orient='records')
print 'Plotting...'
emp_dataframe=empirical_dataframe()
sns.set(context='poster')
figure, (ax1, ax2) = plt.subplots(2, 1)
sns.tsplot(time="time",value="wm",data=primary_dataframe,unit="trial",condition='drug',ax=ax1,ci=95)
sns.tsplot(time="time",value="correct",data=primary_dataframe,unit="trial",condition='drug',ax=ax2,ci=95)
sns.tsplot(time="time",value="accuracy",data=emp_dataframe,unit='trial',condition='drug',
interpolate=False,ax=ax2,color=sns.color_palette('dark'))
sns.tsplot(time="time",value="accuracy",data=emp_dataframe, unit='trial',condition='drug',
interpolate=True,ax=ax2,color=sns.color_palette('dark'),legend=False)
ax1.set(xlabel='',ylabel='decoded $\hat{cue}$',xlim=(0,9.5),ylim=(0,1),
title="drug_type=%s, decision_type=%s, trials=%s" %(drug_type,decision_type,n_trials))
ax2.set(xlabel='time (s)',xlim=(0,9.5),ylim=(0.5,1),ylabel='DRT accuracy')
figure.savefig('primary_plots.png')
figure2, (ax3, ax4) = plt.subplots(1, 2)
if len(firing_dataframe.query("tuning=='weak'"))>0:
sns.tsplot(time="time",value="firing_rate",unit="neuron-trial",condition='drug',ax=ax3,ci=95,
data=firing_dataframe.query("tuning=='weak'").reset_index())
if len(firing_dataframe.query("tuning=='nonpreferred'"))>0:
sns.tsplot(time="time",value="firing_rate",unit="neuron-trial",condition='drug',ax=ax4,ci=95,
data=firing_dataframe.query("tuning=='nonpreferred'").reset_index())
ax3.set(xlabel='time (s)',xlim=(0.0,9.5),ylim=(0,250),ylabel='Normalized Firing Rate',title='Preferred Direction')
ax4.set(xlabel='time (s)',xlim=(0.0,9.5),ylim=(0,250),ylabel='',title='Nonpreferred Direction')
figure2.savefig('firing_plots.png')
