def test_list(self):
#generate a List sampler input
sampler=spotpy.algorithms.mc(spot_setup(),parallel=self.parallel, dbname='Rosen', dbformat='csv', sim_timeout=self.timeout)
sampler.sample(self.rep)
sampler=spotpy.algorithms.list_sampler(spot_setup(),parallel=self.parallel, dbname='Rosen', dbformat=self.dbformat, sim_timeout=self.timeout)
sampler.sample(self.rep)
results = sampler.getdata()
self.assertEqual(len(results), self.rep)
def setup_griewank(self):
if not os.path.isfile("setup_griewank_pickle"):
from spotpy.examples.spot_setup_griewank import spot_setup
# Create samplers for every algorithm:
spot_setup = spot_setup()
rep = 100
timeout = 10 # Given in Seconds
parallel = "seq"
dbformat = "csv"
sampler = spotpy.algorithms.mc(spot_setup,
parallel=parallel,
dbname='RosenMC',
dbformat=dbformat,
sim_timeout=timeout)
sampler.sample(rep)
fl = open("setup_griewank_pickle", "wb")
pickle.dump(
{
"getdata": sampler.getdata(),
"evaluation": sampler.evaluation
}, fl)
fl.close()
with open("setup_griewank_pickle", "rb") as file:
return pickle.load(file)
def test_plot_fast_sensitivity(self):
from spotpy.examples.spot_setup_rosenbrock import spot_setup
spot_setup_object = spot_setup()
parallel = "seq"
dbformat = "ram"
timeout = 5
sampler = spotpy.algorithms.fast(spot_setup_object,
parallel=parallel,
dbname='test_get_sensitivity_of_fast',
dbformat=dbformat,
sim_timeout=timeout)
sampler.sample(300)
results = []
results.append(sampler.getdata())
results = np.array(results)[0]
print("Sampler is done with")
print(results)
spotpy.analyser.plot_fast_sensitivity(results)
fig_name = "FAST_sensitivity.png"
# approximately 8855 KB is the size of an empty matplotlib.pyplot.plot, so
# we expecting a plot with some content without testing the structure of the pot just
# the size
self.assertGreaterEqual(os.path.getsize(fig_name), 8855)
# tidy up all
os.remove(fig_name)
def test_demcz(self):
sampler = spotpy.algorithms.demcz(spot_setup(GausianLike),
parallel=self.parallel,
dbname='Rosen',
dbformat=self.dbformat,
sim_timeout=self.timeout)
sampler.sample(self.rep, convergenceCriteria=0)
results = sampler.getdata()
self.assertEqual(len(results), self.rep)
def test_fscabc(self):
sampler = spotpy.algorithms.fscabc(spot_setup(used_algorithm='fscabc'),
parallel=self.parallel,
dbname='Rosen',
dbformat=self.dbformat,
sim_timeout=self.timeout)
sampler.sample(self.rep)
results = sampler.getdata()
self.assertEqual(len(results), self.rep)
def test_fast(self):
sampler = spotpy.algorithms.fast(spot_setup(),
parallel=self.parallel,
dbname='Rosen',
dbformat=self.dbformat,
sim_timeout=self.timeout)
sampler.sample(self.rep, M=5)
results = sampler.getdata()
self.assertEqual(len(results), self.rep) #Si values should be returned
def test_umpc(self):
sampler = spotpy.algorithms.mc(spot_setup(),
parallel='umpc',
dbname='Rosen',
dbformat=self.dbformat,
sim_timeout=self.timeout)
sampler.sample(self.rep)
results = sampler.getdata()
self.assertEqual(len(results), self.rep)
def setUp(self):
# How many digits to match in case of floating point answers
self.tolerance = 7
#Create samplers for every algorithm:
self.spot_setup = spot_setup()
self.rep = 987
self.timeout = 10 #Given in Seconds
self.parallel = os.environ.get('SPOTPY_PARALLEL', 'seq')
self.dbformat = "ram"
def test_dream(self):
sampler = spotpy.algorithms.dream(spot_setup(GausianLike),
parallel=self.parallel,
dbname='Rosen',
dbformat=self.dbformat,
sim_timeout=self.timeout)
sampler.sample(self.rep,
convergence_limit=0.9,
runs_after_convergence=500)
results = sampler.getdata()
self.assertEqual(len(results), self.rep)
def test_sceua(self):
sampler = spotpy.algorithms.sceua(spot_setup(),
parallel=self.parallel,
dbname='Rosen',
dbformat=self.dbformat,
sim_timeout=self.timeout)
sampler.sample(self.rep)
results = sampler.getdata()
self.assertLessEqual(
len(results),
self.rep) #Sceua save per definition not all sampled runs
def setUp(self):
np.random.seed(42)
self.rep = 100
if not os.path.isfile("setUp_pickle_file"):
from spotpy.examples.spot_setup_rosenbrock import spot_setup
spot_setup_object = spot_setup()
self.results = []
parallel = "seq"
dbformat = "ram"
timeout = 5
self.sampler = spotpy.algorithms.dream(spot_setup_object,
parallel=parallel,
dbname='RosenDREAM',
dbformat=dbformat,
sim_timeout=timeout)
self.sampler.sample(self.rep)
self.results.append(self.sampler.getdata())
self.results = np.array(self.results)
def test_plot_regression(self):
from spotpy.examples.spot_setup_rosenbrock import spot_setup
spot_setup_object = spot_setup()
parallel = "mpc"
dbformat = "ram"
timeout = 5
sampler = spotpy.algorithms.mc(spot_setup_object,
parallel=parallel,
dbname='test_plot_regression',
dbformat=dbformat,
sim_timeout=timeout)
sampler.sample(300)
spotpy.analyser.plot_regression(sampler.getdata(), sampler.evaluation)
fig_name = "regressionanalysis.png"
# approximately 8855 KB is the size of an empty matplotlib.pyplot.plot, so
# we expecting a plot with some content without testing the structure of the pot just
# the size
self.assertGreaterEqual(os.path.getsize(fig_name), 8855)
os.remove(fig_name)
def test_get_sensitivity_of_fast(self):
from spotpy.examples.spot_setup_rosenbrock import spot_setup
spot_setup_object = spot_setup()
results = []
parallel = "seq"
dbformat = "ram"
timeout = 5
self.sampler = spotpy.algorithms.fast(
spot_setup_object,
parallel=parallel,
dbname='test_get_sensitivity_of_fast',
dbformat=dbformat,
sim_timeout=timeout)
self.sampler.sample(200)
results.append(self.sampler.getdata())
results = np.array(results)[0]
get_sensitivity_of_fast = spotpy.analyser.get_sensitivity_of_fast(
results)
self.assertEqual(len(get_sensitivity_of_fast), 2)
self.assertEqual(len(get_sensitivity_of_fast["S1"]), 3)
self.assertEqual(len(get_sensitivity_of_fast["ST"]), 3)
self.assertEqual(type(get_sensitivity_of_fast), type({}))
def test_cli_run(self):
setup = spot_setup()
runner = CliRunner()
result = runner.invoke(run, ['-n 10', '-s', 'lhs', '-p', 'seq'], obj=setup)
self.assertEqual(result.exit_code, 0)
Copyright 2015 by Tobias Houska This file is part of Statistical Parameter Estimation Tool (SPOTPY). :author: Tobias Houska This class holds the example code from the Rosenbrock tutorial web-documention. ''' from __future__ import absolute_import from __future__ import division from __future__ import print_function from __future__ import unicode_literals import spotpy from spotpy.examples.spot_setup_rosenbrock import spot_setup from spotpy.describe import describe #Create samplers for every algorithm: results=[] spot_setup=spot_setup() rep=5000 sampler=spotpy.algorithms.mc(spot_setup, dbname='RosenMC', dbformat='csv') print(describe(sampler)) sampler.sample(rep) results.append(sampler.getdata()) sampler=spotpy.algorithms.lhs(spot_setup, dbname='RosenLHS', dbformat='csv') print(describe(sampler)) sampler.sample(rep) results.append(sampler.getdata()) sampler=spotpy.algorithms.mle(spot_setup, dbname='RosenMLE', dbformat='csv') print(describe(sampler)) sampler.sample(rep)
''' Copyright 2015 by Tobias Houska This file is part of Statistical Parameter Estimation Tool (SPOTPY). :author: Tobias Houska This class holds the example code from the Rosenbrock tutorial web-documention. ''' from __future__ import absolute_import from __future__ import division from __future__ import print_function from __future__ import unicode_literals import spotpy from spotpy.examples.spot_setup_rosenbrock import spot_setup #Create samplers for every algorithm: results=[] spot_setup=spot_setup() rep=5000 sampler=spotpy.algorithms.mc(spot_setup, dbname='RosenMC', dbformat='csv') sampler.sample(rep) results.append(sampler.getdata()) sampler=spotpy.algorithms.lhs(spot_setup, dbname='RosenLHS', dbformat='csv') sampler.sample(rep) results.append(sampler.getdata()) sampler=spotpy.algorithms.mle(spot_setup, dbname='RosenMLE', dbformat='csv') sampler.sample(rep) results.append(sampler.getdata()) sampler=spotpy.algorithms.mcmc(spot_setup, dbname='RosenMCMC', dbformat='csv')
This class holds the example code from the getting_started web-documention.
'''
# Getting started
#To start your experience with SPOT you need to have SPOT installed. Please see the [Installation chapter](index.md) for further details.
#To use SPOT we have to import it and use one of the pre-build examples:
import spotpy # Load the SPOT package into your working storage
from spotpy import analyser # Load the Plotting extension
from spotpy.examples.spot_setup_rosenbrock import spot_setup # Import the two dimensional Rosenbrock example
#from spot_setup_griewank import spot_setup
#The example comes along with parameter boundaries, the Rosenbrock function, the optimal value of the function and RMSE as a likelihood.
#So we can directly start to analyse the Rosenbrock function with one of the algorithms. We start with a simple Monte Carlo sampling:
if __name__ == '__main__':
# Give Monte Carlo algorithm the example setup and saves results in a RosenMC.csv file
#spot_setup.slow = True
sampler = spotpy.algorithms.mc(spot_setup(), dbname='RosenMC', dbformat='ram')
#Now we can sample with the implemented Monte Carlo algortihm:
sampler.sample(100000) # Sample 100.000 parameter combinations
results=sampler.getdata()
#Now we want to have a look at the results. First we want to know, what the algorithm has done during the 10.000 iterations:
#spot.analyser.plot_parametertrace(results) # Use the analyser to show the parameter trace
spotpy.analyser.plot_parameterInteraction(results)
posterior=spotpy.analyser.get_posterior(results)
spotpy.analyser.plot_parameterInteraction(posterior)
#spotpy.analyser.plot_posterior_parametertrace(results, threshold=0.9)
print spotpy.analyser.get_best_parameterset(results)
from spotpy.examples.spot_setup_rosenbrock import spot_setup
from spotpy.describe import describe
#Create samplers for every algorithm:
results = []
rep = 1000
timeout = 10 #Given in Seconds
parallel = "seq"
dbformat = "csv"
# Bayesian algorithms should be run with a likelihood function
bayesian_likelihood_func = spotpy.likelihoods.gaussianLikelihoodMeasErrorOut
sampler = spotpy.algorithms.mc(spot_setup(),
parallel=parallel,
dbname='RosenMC',
dbformat=dbformat,
sim_timeout=timeout)
print(describe(sampler))
sampler.sample(rep)
results.append(sampler.getdata())
sampler = spotpy.algorithms.lhs(spot_setup(),
parallel=parallel,
dbname='RosenLHS',
dbformat=dbformat,
sim_timeout=timeout)
sampler.sample(rep)
results.append(sampler.getdata())
This class holds the example code from the getting_started web-documention.
'''
from __future__ import print_function, division, absolute_import, unicode_literals
# Getting started
#To start your experience with SPOT you need to have SPOT installed. Please see the [Installation chapter](index.md) for further details.
#To use SPOT we have to import it and use one of the pre-build examples:
import spotpy # Load the SPOT package into your working storage
from spotpy.examples.spot_setup_rosenbrock import spot_setup # Import the two dimensional Rosenbrock example
#The example comes along with parameter boundaries, the Rosenbrock function, the optimal value of the function and RMSE as a likelihood.
#So we can directly start to analyse the Rosenbrock function with one of the algorithms. We start with a simple Monte Carlo sampling:
if __name__ == '__main__':
# Give Monte Carlo algorithm the example setup and saves results in a RosenMC.csv file
#spot_setup.slow = True
sampler = spotpy.algorithms.mc(spot_setup(),
dbname='RosenMC',
dbformat='ram')
#Now we can sample with the implemented Monte Carlo algortihm:
sampler.sample(100000) # Sample 100.000 parameter combinations
results = sampler.getdata()
#Now we want to have a look at the results. First we want to know, what the algorithm has done during the 10.000 iterations:
#spot.analyser.plot_parametertrace(results) # Use the analyser to show the parameter trace
spotpy.analyser.plot_parameterInteraction(results)
posterior = spotpy.analyser.get_posterior(results)
spotpy.analyser.plot_parameterInteraction(posterior)
#spotpy.analyser.plot_posterior_parametertrace(results, threshold=0.9)
print(spotpy.analyser.get_best_parameterset(results))
