def performReevaluation(model, params, policies, outputFile):
if (params.createNewReevaluationResults):
with MultiprocessingEvaluator(model) as evaluator:
scenarios = params.evaluationScenarios
results = evaluator.perform_experiments(scenarios=scenarios,
policies=policies)
if not os.path.exists(params.reevaluateOutputFolder):
os.makedirs(params.reevaluateOutputFolder)
save_results(results, params.reevaluateOutputFolder + outputFile)
else:
print('Loading reevaluation from ' + params.reevaluateOutputFolder +
outputFile)
results = load_results(params.reevaluateOutputFolder + outputFile)
return results
def load_ema_leso_results(
run_id: int,
exp_prefix: str,
results_folder: str,
return_db_as_df=True,
exclude_solver_errors=True,
) -> Tuple[pd.DataFrame, dict, pd.DataFrame]:
"""Small helper function to load results easily from the document structure"""
ema_results = f"{exp_prefix}_ema_results_{run_id}.tar.gz"
experiments, outcomes = ema_workbench.load_results(results_folder /
ema_results)
db_file = f"{exp_prefix}_db{run_id}.json"
db = TinyDB(results_folder / db_file)
if return_db_as_df:
db = convert_db_to_df(db)
if exclude_solver_errors:
db = db[db.solver_status == "ok"]
return experiments, outcomes, db
'''
'''
import matplotlib.pyplot as plt
import seaborn as sns
from ema_workbench import load_results
import ema_workbench.analysis.logistic_regression as logistic_regression
# Created on 14 Mar 2019
#
# .. codeauthor:: jhkwakkel <j.h.kwakkel (at) tudelft (dot) nl>
experiments, outcomes = load_results('./data/1000 flu cases no policy.tar.gz')
x = experiments.drop(['model', 'policy'], axis=1)
y = outcomes['deceased population region 1'][:, -1] > 1000000
logit = logistic_regression.Logit(x, y)
logit.run()
logit.show_tradeoff()
# when we change the default threshold, the tradeoff curve is
# recalculated
logit.threshold = 0.8
logit.show_tradeoff()
# we can also look at the tradeoff across threshold values
'''
'''
import matplotlib.pyplot as plt
import seaborn as sns
from ema_workbench import load_results
import ema_workbench.analysis.logistic_regression as logistic_regression
# Created on 14 Mar 2019
#
# .. codeauthor:: jhkwakkel <j.h.kwakkel (at) tudelft (dot) nl>
experiments, outcomes = load_results('./data/1000 flu cases no policy.tar.gz')
x = experiments.drop(['model', 'policy'], axis=1)
y = outcomes['deceased population region 1'][:, -1] > 1000000
logit = logistic_regression.Logit(x, y)
logit.run()
logit.show_tradeoff()
# when we change the default threshold, the tradeoff curve is
# recalculated
logit.threshold = 0.8
logit.show_tradeoff()
# we can also look at the tradeoff across threshold values
# for a given model
logit.show_threshold_tradeoff(3)
scenario_data[i][j] = data[ooi][s]
distances = pdist(scenario_data, distance)
return distances
model = sys.argv[1]
selectType = sys.argv[2]
nr_policies = sys.argv[3]
nr_experiments = sys.argv[4]
set_size = int(sys.argv[5])
dir = '../data/multi/scenarioselection/scens_pol' + nr_policies + '/'
fn = model + '_' + selectType + 'selected_' + nr_experiments + 'experiments_' + nr_policies + 'policies.tar.gz'
print(dir + fn)
try:
results = load_results(dir + fn)
except:
print('skipping results')
exit(0)
exp, outcomes = results
norm_new_out = normalize_out_dic(outcomes)
oois = list(outcomes.keys())
def evaluate_diversity_single(x,
data=norm_new_out,
oois=oois,
weight=0.5,
distance='euclidean'):
'''
takes the outcomes and selected scenario set (decision variables),
.. codeauthor:: jhkwakkel <j.h.kwakkel (at) tudelft (dot) nl>
'''
import numpy as np
import matplotlib.pyplot as plt
from ema_workbench import load_results, ema_logging
from ema_workbench.analysis.pairs_plotting import (pairs_lines, pairs_scatter,
pairs_density)
ema_logging.log_to_stderr(level=ema_logging.DEFAULT_LEVEL)
# load the data
fh = './data/1000 flu cases no policy.tar.gz'
experiments, outcomes = load_results(fh)
# transform the results to the required format
# that is, we want to know the max peak and the casualties at the end of the
# run
tr = {}
# get time and remove it from the dict
time = outcomes.pop('TIME')
for key, value in outcomes.items():
if key == 'deceased population region 1':
tr[key] = value[:, -1] #we want the end value
else:
# we want the maximum value of the peak
max_peak = np.max(value, axis=1)
from __future__ import (absolute_import, print_function, division, \
unicode_literals)
from ema_workbench import load_results
# ####################### IMPORT DATA #########################################
# =============================================================================
results = load_results('./800experiments_300s_7%_ForReal.tar.gz')
experiments, outcomes = results
Created on May 26, 2015 @author: jhkwakkel ''' import matplotlib.pyplot as plt import ema_workbench.analysis.cart as cart from ema_workbench import ema_logging, load_results ema_logging.log_to_stderr(level=ema_logging.INFO) default_flow = 2.178849944502783e7 # load data fn = './data/5000 runs WCM.tar.gz' results = load_results(fn) x, outcomes = results ooi = 'throughput Rotterdam' outcome = outcomes[ooi] / default_flow y = outcome < 1 cart_alg = cart.CART(x, y) cart_alg.build_tree() # print cart to std_out print(cart_alg.stats_to_dataframe()) print(cart_alg.boxes_to_dataframe()) # visualize cart_alg.show_boxes(together=False)
''' This file illustrated the use of the workbench for using dimensional stacking for scenario discovery .. codeauthor:: jhkwakkel <j.h.kwakkel (at) tudelft (dot) nl> ''' import matplotlib.pyplot as plt from ema_workbench import ema_logging, load_results from ema_workbench.analysis import dimensional_stacking ema_logging.log_to_stderr(level=ema_logging.INFO) # load data fn = './data/1000 flu cases no policy.tar.gz' x, outcomes = load_results(fn) y = outcomes['deceased population region 1'][:, -1] > 1000000 fig = dimensional_stacking.create_pivot_plot(x, y, 2, bin_labels=True) plt.show()
""" Created on Jul 8, 2014 @author: [email protected] """ import matplotlib.pyplot as plt from ema_workbench import ema_logging, load_results from ema_workbench.analysis.plotting import envelopes from ema_workbench.analysis.plotting_util import KDE ema_logging.log_to_stderr(ema_logging.INFO) file_name = r"./data/1000 flu cases.tar.gz" results = load_results(file_name) # the plotting functions return the figure and a dict of axes fig, axes = envelopes(results, group_by="policy", density=KDE, fill=True) # we can access each of the axes and make changes for key, value in axes.iteritems(): # the key is the name of the outcome for the normal plot # and the name plus '_density' for the endstate distribution if key.endswith("_density"): value.set_xscale("log") plt.show()
ScalarOutcome('reliability', kind=ScalarOutcome.MAXIMIZE)
]
# override some of the defaults of the model
lake_model.constants = [
Constant('alpha', 0.41),
Constant('nsamples', 100),
Constant('timehorizon', lake_model.time_horizon),
]
#Load the initial exploration for scenarios
#directory = 'D:/sibeleker/surfdrive/Documents/Notebooks/Lake_model-MORDM/Sibel/data/'
#fn = '211_experiments_closedloop_noApollution_inertia.tar.gz'
directory = 'H:/MyDocuments/Notebooks/Lake_model-MORDM/Sibel/MORDM_paper/data/'
#fn = '206_experiments_openloop_Apollution.tar.gz'
fn = '500_experiments_openloop_5policies.tar.gz'
scenario_results = load_results(directory + fn)
experiments, outcomes = scenario_results
#get the selected scenarios
#scenarios = [77, 96, 130, 181]
scenarios = ['Ref', 77, 96, 130, 181]
random_scenarios = [81, 289, 391, 257]
#optimize for each selected scenario
pool = multiprocessing.Pool(processes=8)
timeout = 5000
start = time.time()
for s in random_scenarios:
if s == 'Ref':
result = pool.apply_async(optimize,
args=(lake_model, {}, 10000,
[0.1, 0.05, 0.005,
''' Created on Jul 8, 2014 @author: [email protected] ''' import matplotlib.pyplot as plt from ema_workbench import ema_logging, load_results from ema_workbench.analysis.plotting import envelopes from ema_workbench.analysis.plotting_util import KDE ema_logging.log_to_stderr(ema_logging.INFO) file_name = r'./data/1000 flu cases.tar.gz' results = load_results(file_name) # the plotting functions return the figure and a dict of axes fig, axes = envelopes(results, group_by='policy', density=KDE, fill=True) # we can access each of the axes and make changes for key, value in axes.iteritems(): # the key is the name of the outcome for the normal plot # and the name plus '_density' for the endstate distribution if key.endswith('_density'): value.set_xscale('log') plt.show()
sqeuclidian, wminkowski
returns a list of distance values
'''
#make a matrix of the data n_scenarios x oois
scenario_data = np.zeros((len(scenarios), len(oois)))
for i, s in enumerate(scenarios):
for j, ooi in enumerate(oois):
scenario_data[i][j] = data[ooi][s]
distances = pdist(scenario_data, distance)
return distances
# dir = 'H:/MyDocuments/Notebooks/Lake_model-MORDM/Sibel/MORDM_paper/data/'
# fn = '206_experiments_openloop_Apollution.tar.gz'
results = load_results('./selected_results.tar.gz')
exp, outcomes = results
norm_new_out = normalize_out_dic(outcomes)
oois = list(outcomes.keys())
def evaluate_diversity_single(x,
data=norm_new_out,
oois=oois,
weight=0.5,
distance='euclidean'):
'''
takes the outcomes and selected scenario set (decision variables),
returns a single 'diversity' value for the scenario set.
outcomes : outcomes dictionary of the scenario ensemble
decision vars : indices of the scenario set
.. codeauthor:: jhkwakkel <j.h.kwakkel (at) tudelft (dot) nl>
'''
import numpy as np
import matplotlib.pyplot as plt
from ema_workbench import load_results, ema_logging
from ema_workbench.analysis.pairs_plotting import (pairs_lines, pairs_scatter,
pairs_density)
ema_logging.log_to_stderr(level=ema_logging.DEFAULT_LEVEL)
# load the data
fh = './data/1000 flu cases no policy.tar.gz'
experiments, outcomes = load_results(fh)
# transform the results to the required format
# that is, we want to know the max peak and the casualties at the end of the
# run
tr = {}
# get time and remove it from the dict
time = outcomes.pop('TIME')
for key, value in outcomes.items():
if key == 'deceased population region 1':
tr[key] = value[:,-1] #we want the end value
else:
# we want the maximum value of the peak
max_peak = np.max(value, axis=1)
# model_version = 'v4'
import os
from dest_directories import gz_path, fig_path
#%%
# n_scenarios = 2000
# n_policies = 50
# run = '36_OE'
n_scenarios = 100000
n_policies = 1
run = 'run_35_NordOE'
# %%
results = load_results(
os.path.join(gz_path, 'run_35_NordOE_v7_100000s_.tar.gz'))
#%%
experiments, outcomes = results
#%%
experiments = experiments.drop(['model', 'policy'], axis=1)
# %%
TimeLine = []
for i in range(65):
TimeLine.append(2020 + i * 5)
TimeLine
outcomes["TIME"] = np.array([TimeLine])
# outcomes
# %%
import ema_workbench
import os
from ema_workbench import experiments_to_scenarios
from ema_workbench.em_framework import SequentialEvaluator
import numpy as np
from cablepool_leso_handshake import METRICS
RESULT_FOLDER = "C:\\Users\\sethv\\Google Drive\\0 Thesis\\Results\\cablepooling"
ema_results = "cabelpooling_ema_results_120821.tar.gz"
ema_results_path = os.path.join(RESULT_FOLDER, ema_results)
experiments, outcomes = ema_workbench.load_results(ema_results_path)
scenarios = experiments_to_scenarios(experiments)
from cablepool_ema_model_definition import model
def printer(**kwargs):
# for key, value in kwargs.items():
# print(f"{key}: {value}")
return {key: np.random.random() for key in METRICS}
model.function = printer
with SequentialEvaluator(model) as evaluator:
results = evaluator.perform_experiments(scenarios=scenarios)
@author: [email protected] ''' import matplotlib.pyplot as plt import numpy as np from ema_workbench import ema_logging, load_results from ema_workbench.analysis.plotting import envelopes, lines from ema_workbench.analysis.plotting_util import KDE ema_logging.log_to_stderr(ema_logging.INFO) file_name = r'./data/10 runs.tar.gz' #file_name = r'./data/1000 flu cases no policy.tar.gz' experiments, outcomes = load_results(file_name) results = (experiments, outcomes) default_flow = 2.178849944502783e7 ooi_name = "sheep" outcome = outcomes[ooi_name] outcome = outcome/default_flow ooi = np.zeros(outcome.shape[0]) temp_outcomes = {ooi_name: ooi} print ooi.shape desired__nr_lines = 5 nr_cases = ooi.shape[0]
'''
Created on 30 Oct 2018
@author: jhkwakkel
'''
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
from ema_workbench import ema_logging, load_results
from ema_workbench.analysis import feature_scoring
ema_logging.log_to_stderr(level=ema_logging.INFO)
# load data
fn = r'./data/1000 flu cases with policies.tar.gz'
x, outcomes = load_results(fn)
# we have timeseries so we need scalars
y = {
'deceased population': outcomes['deceased population region 1'][:, -1],
'max. infected fraction': np.max(outcomes['infected fraction R1'], axis=1)
}
scores = feature_scoring.get_feature_scores_all(x, y)
sns.heatmap(scores, annot=True, cmap='viridis')
plt.show()
''' Created on Jul 8, 2014 @author: [email protected] ''' import matplotlib.pyplot as plt from ema_workbench import ema_logging, load_results from ema_workbench.analysis.plotting import kde_over_time ema_logging.log_to_stderr(ema_logging.INFO) # file_name = r'./data/1000 runs scarcity.tar.gz' file_name = './data/1000 flu cases no policy.tar.gz' experiments, outcomes = load_results(file_name) # the plotting functions return the figure and a dict of axes fig, axes = kde_over_time(experiments, outcomes, log=True) plt.show()
def classify(data):
# get the output for deceased population
result = data['deceased population region 1']
# if deceased population is higher then 1.000.000 people,
# classify as 1
classes = result[:, -1] > 1000000
return classes
# load data
fn = './data/1000 flu cases with policies.tar.gz'
results = load_results(fn)
experiments, results = results
# extract results for 1 policy
logical = experiments['policy'] == 'no policy'
new_experiments = experiments[logical]
new_results = {}
for key, value in results.items():
new_results[key] = value[logical]
results = (new_experiments, new_results)
# perform cart on modified results tuple
cart_alg = cart.setup_cart(results, classify, mass_min=0.05)
cart_alg.build_tree()
Created on Jul 8, 2014 @author: [email protected] ''' import matplotlib.pyplot as plt import numpy as np from ema_workbench import ema_logging, load_results from ema_workbench.analysis.plotting import envelopes, lines from ema_workbench.analysis.plotting_util import KDE ema_logging.log_to_stderr(ema_logging.INFO) experiments, outcomes = load_results(r'./Data/500_runs_cap_approach.tar.gz') #results = (experiments, outcomes) #experiments, outcomes = results print(experiments.shape) #500 scenarios print(outcomes['level-of-achievability'].shape) fig, ax = plt.subplots(1) ax.plot(outcomes['level-of-achievability'][:,0,:].T) ax.set_ylabel('Percentage') fig.set_size_inches(6,6) plt.show()