CUQIpy 1.1.0.post0.dev47__tar.gz → 1.1.1.post0.dev7__tar.gz

{cuqipy-1.1.0.post0.dev47 → cuqipy-1.1.1.post0.dev7}/CUQIpy.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: CUQIpy
-Version: 1.1.0.post0.dev47
+Version: 1.1.1.post0.dev7
 Summary: Computational Uncertainty Quantification for Inverse problems in Python
 Maintainer-email: "Nicolai A. B. Riis" <nabr@dtu.dk>, "Jakob S. Jørgensen" <jakj@dtu.dk>, "Amal M. Alghamdi" <amaal@dtu.dk>, Chao Zhang <chaz@dtu.dk>
 License:                                  Apache License

{cuqipy-1.1.0.post0.dev47 → cuqipy-1.1.1.post0.dev7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: CUQIpy
-Version: 1.1.0.post0.dev47
+Version: 1.1.1.post0.dev7
 Summary: Computational Uncertainty Quantification for Inverse problems in Python
 Maintainer-email: "Nicolai A. B. Riis" <nabr@dtu.dk>, "Jakob S. Jørgensen" <jakj@dtu.dk>, "Amal M. Alghamdi" <amaal@dtu.dk>, Chao Zhang <chaz@dtu.dk>
 License:                                  Apache License

{cuqipy-1.1.0.post0.dev47 → cuqipy-1.1.1.post0.dev7}/cuqi/_version.py

@@ -8,11 +8,11 @@ import json
 
 version_json = '''
 {
- "date": "2024-08-29T13:49:05+0200",
+ "date": "2024-09-03T13:01:53+0200",
  "dirty": false,
  "error": null,
- "full-revisionid": "67dba338b95cc7c16357687a686de0a3c3e5b512",
- "version": "1.1.0.post0.dev47"
+ "full-revisionid": "6731d720ebd90ec077628febef51db45ce1738a4",
+ "version": "1.1.1.post0.dev7"
 }
 '''  # END VERSION_JSON
 

{cuqipy-1.1.0.post0.dev47 → cuqipy-1.1.1.post0.dev7}/cuqi/distribution/_joint_distribution.py

@@ -175,6 +175,10 @@ class JointDistribution:
                 kwargs[ordered_keys[index]] = arg
         return kwargs
 
+    def _sum_evaluated_densities(self):
+        """ Return the sum of the evaluated densities in the joint distribution """
+        return sum([density.logd() for density in self._evaluated_densities])
+
     def _reduce_to_single_density(self):
         """ Reduce the joint distribution to a single density if possible.
 
@@ -183,7 +187,6 @@ class JointDistribution:
         This method is a hack to allow our current samplers to work with
         the joint distribution. It should be removed in the future.
         """
-
         # Count number of distributions and likelihoods
         n_dist = len(self._distributions)
         n_likelihood = len(self._likelihoods)
@@ -201,16 +204,29 @@ class JointDistribution:
             # Ensure parameter names match, otherwise return the joint distribution
             if set(self._likelihoods[0].get_parameter_names()) != set(self._distributions[0].get_parameter_names()):
                 return self
-            return Posterior(self._likelihoods[0], self._distributions[0])
-        
+            return self._add_constants_to_density(Posterior(self._likelihoods[0], self._distributions[0]))
+
         # If exactly one distribution and no likelihoods its a Distribution
         if n_dist == 1 and n_likelihood == 0:
-            return self._distributions[0]
-
+            return self._add_constants_to_density(self._distributions[0])        
+        
         # If no distributions and exactly one likelihood its a Likelihood
         if n_likelihood == 1 and n_dist == 0:
             return self._likelihoods[0]
 
+        # If only evaluated densities left return joint to ensure logd method is available
+        if n_dist == 0 and n_likelihood == 0:
+            return self
+        
+    def _add_constants_to_density(self, density: Density):
+        """ Add the constants (evaluated densities) to a single density. Used when reducing to single density. """
+
+        if isinstance(density, EvaluatedDensity):
+            raise ValueError("Cannot add the sum of all evaluated densities to an EvaluatedDensity.")
+
+        density._constant += self._sum_evaluated_densities()
+        return density
+
     def _as_stacked(self) -> _StackedJointDistribution:
         """ Return a stacked JointDistribution with the same densities. """
         return _StackedJointDistribution(*self._densities)

{cuqipy-1.1.0.post0.dev47 → cuqipy-1.1.1.post0.dev7}/cuqi/experimental/mcmc/_gibbs.py

@@ -6,10 +6,10 @@ import numpy as np
 import warnings
 
 try:
-    from progressbar import progressbar
+    from tqdm import tqdm
 except ImportError:
-    def progressbar(iterable, **kwargs):
-        warnings.warn("Module mcmc: Progressbar not found. Install progressbar2 to get sampling progress.")
+    def tqdm(iterable, **kwargs):
+        warnings.warn("Module mcmc: tqdm not found. Install tqdm to get sampling progress.")
         return iterable
 
 # Not subclassed from Sampler as Gibbs handles multiple samplers and samples multiple parameters
@@ -152,13 +152,13 @@ class HybridGibbs:
 
     def sample(self, Ns) -> 'HybridGibbs':
         """ Sample from the joint distribution using Gibbs sampling """
-        for _ in progressbar(range(Ns)):
+        for _ in tqdm(range(Ns)):
             self.step()
             self._store_samples()
 
     def warmup(self, Nb) -> 'HybridGibbs':
         """ Warmup (tune) the Gibbs sampler """
-        for idx in progressbar(range(Nb)):
+        for idx in tqdm(range(Nb)):
             self.step()
             self.tune(idx)
             self._store_samples()

{cuqipy-1.1.0.post0.dev47 → cuqipy-1.1.1.post0.dev7}/cuqi/experimental/mcmc/_sampler.py

@@ -7,10 +7,10 @@ import cuqi
 from cuqi.samples import Samples
 
 try:
-    from progressbar import progressbar
+    from tqdm import tqdm
 except ImportError:
-    def progressbar(iterable, **kwargs):
-        warnings.warn("Module mcmc: Progressbar not found. Install progressbar2 to get sampling progress.")
+    def tqdm(iterable, **kwargs):
+        warnings.warn("Module mcmc: tqdm not found. Install tqdm to get sampling progress.")
         return iterable
 
 class Sampler(ABC):
@@ -220,7 +220,7 @@ class Sampler(ABC):
         if hasattr(self, "_pre_sample"): self._pre_sample()
 
         # Draw samples
-        for _ in progressbar( range(Ns) ):
+        for _ in tqdm( range(Ns) ):
             
             # Perform one step of the sampler
             acc = self.step()
@@ -260,7 +260,7 @@ class Sampler(ABC):
         if hasattr(self, "_pre_warmup"): self._pre_warmup()
 
         # Draw warmup samples with tuning
-        for idx in progressbar(range(Nb)):
+        for idx in tqdm(range(Nb)):
 
             # Perform one step of the sampler
             acc = self.step()

{cuqipy-1.1.0.post0.dev47 → cuqipy-1.1.1.post0.dev7}/tests/test_joint_distribution.py

@@ -310,3 +310,86 @@ def test_MultipleLikelihoodPosterior_should_raise_if_names_do_not_match():
     with pytest.raises(ValueError, match=r"the same parameter name"):
         cuqi.distribution.MultipleLikelihoodPosterior(y.to_likelihood(1), x, z)
 
+@pytest.mark.parametrize("joint, variables", [
+    (
+        cuqi.distribution.JointDistribution(
+            cuqi.distribution.Normal(0, 1, name="x"),
+            cuqi.distribution.Normal(0, 1, name="y")
+        ),
+        {
+            "x": 0,
+            "y": 0
+        }
+    ),
+    (
+        cuqi.distribution.JointDistribution(
+            cuqi.distribution.Uniform(0, 10, name="d"),
+            cuqi.distribution.Uniform(0, 5, name="s"),
+            cuqi.distribution.Gaussian(np.zeros(8), lambda d: d, name="x"),
+            cuqi.distribution.Gaussian(
+                mean=cuqi.testproblem.Deconvolution1D(dim=8).model,
+                cov=lambda s: s,
+                name="y"
+            )
+        ),
+        {
+            "d": 5,
+            "s": 2,
+            "x": np.zeros(8),
+            "y": cuqi.testproblem.Deconvolution1D(dim=8).data
+        }
+    ),
+    (
+        cuqi.distribution.JointDistribution(
+            cuqi.distribution.Uniform(0, 10, name="d"),
+            cuqi.distribution.Uniform(0, 5, name="s"),
+            cuqi.distribution.Gaussian(np.zeros(8), lambda d: d, name="x"),
+            cuqi.distribution.Gaussian(
+                mean=cuqi.testproblem.Deconvolution1D(dim=8).model,
+                cov=lambda s: s,
+                name="y"
+            )
+        ),
+        {
+            "d": 11, # Out of bounds (all should return -Inf)
+            "s": 2,
+            "x": np.zeros(8),
+            "y": cuqi.testproblem.Deconvolution1D(dim=8).data
+        }
+    )
+    ]
+)
+def test_logd_consistency_when_conditioning(joint, variables):
+    """ Test consistency of logd value when conditioning the joint distribution.
+     
+    This ensures we always return the correct value for logd even when reducing to single density.
+    
+    """
+
+    # True value of logd by fully evaluating the joint distribution
+    true_value = joint.logd(**variables)
+
+    # Loop over all variables and evaluate the logd value
+    # where all previously seen variables are conditioned
+    # and not seen variables are given to logd.
+    cond_vars = {}
+    for key, value in variables.items():
+        # Remaining variables are those that are not used for conditioning
+        remaining_vars = {k: v for k, v in variables.items() if k not in cond_vars}
+        
+        # Condition the joint distribution
+        cond_joint = joint(**cond_vars)
+
+        # Evaluate the logd value of conditioned joint distribution
+        # This may be a single density or a joint distribution since
+        # joint can reduce to single density (such as Posterior)
+        logd_value = cond_joint.logd(**remaining_vars)
+
+        # Potential error message if this assert fails:
+        msg = f"Evaluated: \n {joint} \n with variables {variables}.\nFailed at {key}={value} for \n {cond_joint} \n with variables {remaining_vars}."
+        
+        # Assert the logd value matches
+        assert logd_value == pytest.approx(true_value, rel=1e-6), msg
+
+        # Add current variable to the variables that need to be conditioned
+        cond_vars[key] = value