Merge pull request #338 from pybop-team/gauss-log-like-fixes

Updates to likelihoods, bug fix `GaussianLogLikelihood` class

CHANGELOG.md

@@ -25,6 +25,7 @@
 
 ## Bug Fixes
 
+- [#338](https://github.com/pybop-team/PyBOP/pull/338) - Fixes GaussianLogLikelihood class, adds integration tests, updates non-bounded parameter implementation by applying bounds from priors and `boundary_multiplier` argument. Bugfixes to CMAES construction.
 - [#339](https://github.com/pybop-team/PyBOP/issues/339) - Updates the calculation of the cyclable lithium capacity in the spme_max_energy example.
 - [#387](https://github.com/pybop-team/PyBOP/issues/387) - Adds keys to ParameterSet and updates ECM OCV check.
 - [#380](https://github.com/pybop-team/PyBOP/pull/380) - Restore self._boundaries construction for `pybop.PSO`

examples/scripts/spm_MAP.py

@@ -2,8 +2,16 @@
 
 import pybop
 
-# Define model
+# Construct and update initial parameter values
 parameter_set = pybop.ParameterSet.pybamm("Chen2020")
+parameter_set.update(
+    {
+        "Negative electrode active material volume fraction": 0.63,
+        "Positive electrode active material volume fraction": 0.51,
+    }
+)
+
+# Define model
 model = pybop.lithium_ion.SPM(parameter_set=parameter_set)
 
 # Fitting parameters
@@ -12,21 +20,16 @@
         "Negative electrode active material volume fraction",
         prior=pybop.Gaussian(0.6, 0.05),
         bounds=[0.5, 0.8],
+        true_value=parameter_set["Negative electrode active material volume fraction"],
     ),
     pybop.Parameter(
         "Positive electrode active material volume fraction",
         prior=pybop.Gaussian(0.48, 0.05),
         bounds=[0.4, 0.7],
+        true_value=parameter_set["Positive electrode active material volume fraction"],
     ),
 )
 
-# Set initial parameter values
-parameter_set.update(
-    {
-        "Negative electrode active material volume fraction": 0.63,
-        "Positive electrode active material volume fraction": 0.51,
-    }
-)
 # Generate data
 sigma = 0.005
 t_eval = np.arange(0, 900, 3)
@@ -44,8 +47,8 @@
 
 # Generate problem, cost function, and optimisation class
 problem = pybop.FittingProblem(model, parameters, dataset)
-cost = pybop.MAP(problem, pybop.GaussianLogLikelihoodKnownSigma)
-optim = pybop.CMAES(
+cost = pybop.MAP(problem, pybop.GaussianLogLikelihoodKnownSigma, sigma0=sigma)
+optim = pybop.AdamW(
     cost,
     max_unchanged_iterations=20,
     min_iterations=20,
@@ -54,6 +57,7 @@
 
 # Run the optimisation
 x, final_cost = optim.run()
+print("True parameters:", parameters.true_value())
 print("Estimated parameters:", x)
 
 # Plot the timeseries output

examples/scripts/spm_MLE.py

@@ -16,7 +16,6 @@
     pybop.Parameter(
         "Positive electrode active material volume fraction",
         prior=pybop.Gaussian(0.48, 0.05),
-        bounds=[0.4, 0.7],
     ),
 )
 
@@ -44,8 +43,8 @@
 
 # Generate problem, cost function, and optimisation class
 problem = pybop.FittingProblem(model, parameters, dataset)
-likelihood = pybop.GaussianLogLikelihoodKnownSigma(problem, sigma=[0.03, 0.03])
-optim = pybop.CMAES(
+likelihood = pybop.GaussianLogLikelihoodKnownSigma(problem, sigma0=sigma)
+optim = pybop.IRPropMin(
     likelihood,
     max_unchanged_iterations=20,
     min_iterations=20,
@@ -57,7 +56,7 @@
 print("Estimated parameters:", x)
 
 # Plot the timeseries output
-pybop.quick_plot(problem, problem_inputs=x[0:2], title="Optimised Comparison")
+pybop.quick_plot(problem, problem_inputs=x, title="Optimised Comparison")
 
 # Plot convergence
 pybop.plot_convergence(optim)

pybop/__init__.py

@@ -86,7 +86,6 @@
     RootMeanSquaredError,
     SumSquaredError,
     ObserverCost,
-    MAP,
 )
 from .costs.design_costs import (
     DesignCost,
@@ -97,6 +96,7 @@
     BaseLikelihood,
     GaussianLogLikelihood,
     GaussianLogLikelihoodKnownSigma,
+    MAP,
 )
 
 #