diff --git a/examples/OHagan-function/test_OHagan.py b/examples/OHagan-function/test_OHagan.py
index 4715ecc525c2a016e078753b63016e8705a77888..2fc629d3423d7ef3312e0182a8f15d1f0fd18ba0 100644
--- a/examples/OHagan-function/test_OHagan.py
+++ b/examples/OHagan-function/test_OHagan.py
@@ -1,7 +1,8 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
-This test deals with the surrogate modeling of a Ishigami function.
+This test deals with the surrogate modeling of O'Hagan function with 15
+parameters.
You will see how to:
Check the quality of your regression model
@@ -30,6 +31,7 @@ sys.path.append("../../src/bayesvalidrox/")
from pylink.pylink import PyLinkForwardModel
from surrogate_models.inputs import Input
from surrogate_models.surrogate_models import MetaModel
+from surrogate_models.meta_model_engine import MetaModelEngine
from post_processing.post_processing import PostProcessing
from bayes_inference.bayes_inference import BayesInference
from bayes_inference.discrepancy import Discrepancy
@@ -87,11 +89,6 @@ if __name__ == "__main__":
# 9)EBL: Emperical Bayesian Learning
MetaModelOpts.pce_reg_method = 'FastARD'
- # Bootstraping
- # 1) normal 2) fast
- MetaModelOpts.bootstrap_method = 'fast'
- MetaModelOpts.n_bootstrap_itrs = 1#00
-
# Specify the max degree to be compared by the adaptive algorithm:
# The degree with the lowest Leave-One-Out cross-validation (LOO)
# error (or the highest score=1-LOO)estimator is chosen as the final
@@ -99,7 +96,7 @@ if __name__ == "__main__":
MetaModelOpts.pce_deg = 7
# q-quasi-norm 0<q<1 (default=1)
- MetaModelOpts.pce_q_norm = 0.5
+ MetaModelOpts.pce_q_norm = 0.65
# Print summary of the regression results
# MetaModelOpts.verbose = True
@@ -116,10 +113,10 @@ if __name__ == "__main__":
# Sampling methods
# 1) random 2) latin_hypercube 3) sobol 4) halton 5) hammersley 6) korobov
# 7) chebyshev(FT) 8) grid(FT) 9) nested_grid(FT) 10)user
- MetaModelOpts.ExpDesign.sampling_method = 'user'
+ MetaModelOpts.ExpDesign.sampling_method = 'latin_hypercube'
# Provide the experimental design object with a hdf5 file
- MetaModelOpts.ExpDesign.hdf5_file = 'ExpDesign_OHagan_orig.hdf5'
+ # MetaModelOpts.ExpDesign.hdf5_file = 'ExpDesign_OHagan_orig.hdf5'
# Sequential experimental design (needed only for sequential ExpDesign)
MetaModelOpts.ExpDesign.n_new_samples = 1
@@ -166,9 +163,6 @@ if __name__ == "__main__":
'Z': np.load("data/valid_outputs.npy")
}
# >>>>>>>>>>>>>>>>>>>>>> Build Surrogate <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
- # Adaptive sparse arbitrary polynomial chaos expansion
- # PCEModel = MetaModelOpts.create_metamodel()
- from surrogate_models.meta_model_engine import MetaModelEngine
meta_model_engine = MetaModelEngine(MetaModelOpts)
meta_model_engine.run()
PCEModel = meta_model_engine.MetaModel
diff --git a/examples/borehole/borehole.py b/examples/borehole/borehole.py
index 43dd71f21904a7ca7bec4e45a48fd3a37cc33879..b2a0dcd4a08091e05b74cceae3eda3709308ae49 100644
--- a/examples/borehole/borehole.py
+++ b/examples/borehole/borehole.py
@@ -1,9 +1,16 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
-Created on Mon Nov 19 08:56:21 2018
+Author: Farid Mohammadi, M.Sc.
+E-Mail: farid.mohammadi@iws.uni-stuttgart.de
+Department of Hydromechanics and Modelling of Hydrosystems (LH2)
+Institute for Modelling Hydraulic and Environmental Systems (IWS), University
+of Stuttgart, www.iws.uni-stuttgart.de/lh2/
+Pfaffenwaldring 61
+70569 Stuttgart
+
+Created on Mon Sep 12 2022
-@author: farid
"""
import numpy as np
diff --git a/examples/borehole/test_borehole.py b/examples/borehole/test_borehole.py
index 3097b78bc8fd221d18f254b2426b9fdff00987e8..e28e9a30f76afa9deb158e23532af12623f2d8e2 100644
--- a/examples/borehole/test_borehole.py
+++ b/examples/borehole/test_borehole.py
@@ -1,7 +1,7 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
-This test deals with the surrogate modeling of a Ishigami function.
+This test deals with the surrogate modeling of a Borehole function.
You will see how to:
Check the quality of your regression model
@@ -15,7 +15,7 @@ of Stuttgart, www.iws.uni-stuttgart.de/lh2/
Pfaffenwaldring 61
70569 Stuttgart
-Created on Wed Jul 10 2019
+Created on Sep 12 2022
"""
@@ -33,6 +33,8 @@ from surrogate_models.surrogate_models import MetaModel
from post_processing.post_processing import PostProcessing
from bayes_inference.bayes_inference import BayesInference
from bayes_inference.discrepancy import Discrepancy
+from surrogate_models.meta_model_engine import MetaModelEngine
+
import matplotlib
matplotlib.use('agg')
@@ -127,11 +129,6 @@ if __name__ == "__main__":
# 9)EBL: Emperical Bayesian Learning
MetaModelOpts.pce_reg_method = 'OMP'
- # Bootstraping
- # 1) normal 2) fast
- MetaModelOpts.bootstrap_method = 'fast'
- MetaModelOpts.n_bootstrap_itrs = 1#00
-
# Specify the max degree to be compared by the adaptive algorithm:
# The degree with the lowest Leave-One-Out cross-validation (LOO)
# error (or the highest score=1-LOO)estimator is chosen as the final
@@ -163,7 +160,7 @@ if __name__ == "__main__":
# Sequential experimental design (needed only for sequential ExpDesign)
MetaModelOpts.ExpDesign.n_new_samples = 1
- MetaModelOpts.ExpDesign.n_max_samples = 300 # 150
+ MetaModelOpts.ExpDesign.n_max_samples = 300
MetaModelOpts.ExpDesign.mod_LOO_threshold = 1e-16
# ------------------------------------------------
@@ -206,9 +203,7 @@ if __name__ == "__main__":
'flow rate [m$^3$/yr]': np.load("data/valid_outputs.npy")
}
# >>>>>>>>>>>>>>>>>>>>>> Build Surrogate <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
- # Adaptive sparse arbitrary polynomial chaos expansion
- # PCEModel = MetaModelOpts.create_metamodel()
- from surrogate_models.meta_model_engine import MetaModelEngine
+ # Use MetaModelEngine for sequential experimental design
meta_model_engine = MetaModelEngine(MetaModelOpts)
meta_model_engine.run()
PCEModel = meta_model_engine.MetaModel
diff --git a/examples/ishigami/test_ishigami.py b/examples/ishigami/test_ishigami.py
index c7db9c4bfb8fe94cb194043836138162624949e2..313ad38280acfe6bf3a4b8c29b1f94d74da399b7 100644
--- a/examples/ishigami/test_ishigami.py
+++ b/examples/ishigami/test_ishigami.py
@@ -30,6 +30,7 @@ sys.path.append("../../src/bayesvalidrox/")
from pylink.pylink import PyLinkForwardModel
from surrogate_models.inputs import Input
from surrogate_models.surrogate_models import MetaModel
+from surrogate_models.meta_model_engine import MetaModelEngine
from post_processing.post_processing import PostProcessing
from bayes_inference.bayes_inference import BayesInference
from bayes_inference.discrepancy import Discrepancy
@@ -94,11 +95,6 @@ if __name__ == "__main__":
# 9)EBL: Emperical Bayesian Learning
MetaModelOpts.pce_reg_method = 'BCS'
- # Bootstraping
- # 1) normal 2) fast
- MetaModelOpts.bootstrap_method = 'fast'
- MetaModelOpts.n_bootstrap_itrs = 100
-
# Specify the max degree to be compared by the adaptive algorithm:
# The degree with the lowest Leave-One-Out cross-validation (LOO)
# error (or the highest score=1-LOO)estimator is chosen as the final
@@ -117,7 +113,7 @@ if __name__ == "__main__":
MetaModelOpts.add_ExpDesign()
# One-shot (normal) or Sequential Adaptive (sequential) Design
- MetaModelOpts.ExpDesign.method = 'sequential'
+ MetaModelOpts.ExpDesign.method = 'normal'
MetaModelOpts.ExpDesign.n_init_samples = 50
# Sampling methods
@@ -171,9 +167,6 @@ if __name__ == "__main__":
MetaModelOpts.valid_samples = np.load("data/valid_samples.npy")
MetaModelOpts.valid_model_runs = {'Z': np.load("data/valid_outputs.npy")}
# >>>>>>>>>>>>>>>>>>>>>> Build Surrogate <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
- # Adaptive sparse arbitrary polynomial chaos expansion
- # PCEModel = MetaModelOpts.create_metamodel()
- from surrogate_models.meta_model_engine import MetaModelEngine
meta_model_engine = MetaModelEngine(MetaModelOpts)
meta_model_engine.run()
PCEModel = meta_model_engine.MetaModel
diff --git a/examples/model-comparison/test_model_comparison.py b/examples/model-comparison/test_model_comparison.py
index b353cca28ce153a1c5f738f83e710a2a1dc88a08..d5f523871d99b5115437c2f66e8a1ce2a845f503 100644
--- a/examples/model-comparison/test_model_comparison.py
+++ b/examples/model-comparison/test_model_comparison.py
@@ -254,22 +254,7 @@ if __name__ == "__main__":
"cosine": NL4_MetaModel
}
- # Option I: MCMC inference method
- mcmc_params = {
- 'n_steps': 1e5,
- 'n_walkers': 30,
- 'multiprocessing': False,
- 'verbose': False
- }
- opts_mcmc = {
- "inference_method": "MCMC",
- "mcmc_params": mcmc_params,
- "Discrepancy": DiscrepancyOpts,
- "emulator": True,
- "plot_post_pred": True
- }
-
- # Option II: BME Bootstrap
+ # BME Bootstrap optuions
opts_bootstrap = {
"bootstrap": True,
"n_samples": 10000,
diff --git a/examples/pollution/test_pollution.py b/examples/pollution/test_pollution.py
index cbb897481121a75c8bb7d604abd6e8e3cbff5036..f2de52e2caf0196d8d05a2035950e2f28ff8fc70 100644
--- a/examples/pollution/test_pollution.py
+++ b/examples/pollution/test_pollution.py
@@ -21,7 +21,6 @@ import numpy as np
import pandas as pd
import joblib
-
# import bayesvalidrox
# Add BayesValidRox path
import sys
@@ -142,7 +141,7 @@ if __name__ == "__main__":
# One-shot (normal) or Sequential Adaptive (sequential) Design
MetaModelOpts.ExpDesign.method = 'normal'
- MetaModelOpts.ExpDesign.n_init_samples = 150 # 5*ndim
+ MetaModelOpts.ExpDesign.n_init_samples = 150
# Sampling methods
# 1) random 2) latin_hypercube 3) sobol 4) halton 5) hammersley 6) korobov
diff --git a/examples/pollution/test_valid_pollution.py b/examples/pollution/test_valid_pollution.py
index a19e7b6a0f81b54e17e51f7bcfa5d8adfdd7dde0..e2aacbe5d2186c38287e552ae588332150dd519b 100644
--- a/examples/pollution/test_valid_pollution.py
+++ b/examples/pollution/test_valid_pollution.py
@@ -18,10 +18,8 @@ Created on Fri Aug 9 2019
"""
import numpy as np
-import pandas as pd
import joblib
-
# import bayesvalidrox
# Add BayesValidRox path
import sys
@@ -129,7 +127,7 @@ if __name__ == "__main__":
# The degree with the lowest Leave-One-Out cross-validation (LOO)
# error (or the highest score=1-LOO)estimator is chosen as the final
# metamodel. pce_deg accepts degree as a scalar or a range.
- MetaModelOpts.pce_deg = np.arange(6)
+ MetaModelOpts.pce_deg = 6
# q-quasi-norm 0<q<1 (default=1)
# MetaModelOpts.pce_q_norm = 0.75
@@ -142,7 +140,7 @@ if __name__ == "__main__":
# One-shot (normal) or Sequential Adaptive (sequential) Design
MetaModelOpts.ExpDesign.method = 'normal'
- MetaModelOpts.ExpDesign.n_init_samples = 150 # 5*ndim
+ MetaModelOpts.ExpDesign.n_init_samples = 150
# Sampling methods
# 1) random 2) latin_hypercube 3) sobol 4) halton 5) hammersley 6) korobov