Update Aesara version and unpin numpy

conda-envs/environment-dev-py37.yml

@@ -5,15 +5,15 @@ channels:
 - defaults
 dependencies:
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.4
 - cachetools>=4.2.1
 - cloudpickle
 - fastprogress>=0.2.0
 - h5py>=2.7
 - ipython
 - myst-nb
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15
 - numpydoc>=0.9
 - pandas>=0.24
 - pre-commit>=2.8.0

conda-envs/environment-dev-py38.yml

@@ -5,15 +5,15 @@ channels:
 - defaults
 dependencies:
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.4
 - cachetools>=4.2.1
 - cloudpickle
 - fastprogress>=0.2.0
 - h5py>=2.7
 - ipython
 - myst-nb
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15.0
 - numpydoc>=0.9
 - pandas>=0.24.0
 - pre-commit>=2.8.0

conda-envs/environment-dev-py39.yml

@@ -5,15 +5,15 @@ channels:
 - defaults
 dependencies:
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.4
 - cachetools>=4.2.1
 - cloudpickle
 - fastprogress>=0.2.0
 - h5py>=2.7
 - ipython
 - myst-nb
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15.0
 - numpydoc>=0.9
 - pandas
 - pre-commit>=2.8.0

conda-envs/environment-test-py37.yml

@@ -5,7 +5,7 @@ channels:
 - defaults
 dependencies:
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.4
 - cachetools>=4.2.1
 - cloudpickle
@@ -14,7 +14,7 @@ dependencies:
 - ipython
 - libblas=*=*mkl
 - mkl-service
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15.0
 - pandas>=0.24
 - pre-commit>=2.8.0
 - pytest-cov>=2.5

conda-envs/environment-test-py38.yml

@@ -5,7 +5,7 @@ channels:
 - defaults
 dependencies:
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.4
 - cachetools>=4.2.1
 - cloudpickle
@@ -14,7 +14,7 @@ dependencies:
 - ipython
 - libblas=*=*mkl
 - mkl-service
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15.0
 - pandas
 - pre-commit>=2.8.0
 - pytest-cov>=2.5

conda-envs/environment-test-py39.yml

@@ -5,7 +5,7 @@ channels:
 - defaults
 dependencies:
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.4
 - cachetools
 - cloudpickle
@@ -14,7 +14,7 @@ dependencies:
 - ipython>=7.16
 - libblas=*=*mkl
 - mkl-service
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15.0
 - pandas
 - pre-commit>=2.8.0
 - pytest-cov>=2.5

conda-envs/windows-environment-dev-py38.yml

@@ -5,13 +5,13 @@ channels:
 dependencies:
  # base dependencies (see install guide for Windows)
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.4
 - cachetools>=4.2.1
 - cloudpickle
 - fastprogress>=0.2.0
 - h5py
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15.0
 - pandas
 - pip
 - python=3.8

conda-envs/windows-environment-test-py38.yml

@@ -5,7 +5,7 @@ channels:
 dependencies:
  # base dependencies (see install guide for Windows)
 - aeppl=0.0.18
-- aesara=2.3.2
+- aesara=2.3.6
 - arviz>=0.11.2
 - cachetools
 - cloudpickle
@@ -15,7 +15,7 @@ dependencies:
 - mkl==2020.4
 - mkl-service==2.3.0
 - m2w64-toolchain
-- numpy>=1.15.0,<1.22.0
+- numpy>=1.15.0
 - pandas
 - pip
 - python=3.8

pymc/distributions/multivariate.py

@@ -1742,7 +1742,7 @@ def dist(
         dist_shape = (rowchol_cov.shape[0], colchol_cov.shape[0])
 
         # Broadcasting mu
-        mu = at.extra_ops.broadcast_to(a=mu, shape=dist_shape)
+        mu = at.extra_ops.broadcast_to(mu, shape=dist_shape)
 
         mu = at.as_tensor_variable(floatX(mu))
         # mean = median = mode = mu

pymc/ode/ode.py

@@ -158,7 +158,7 @@ def __call__(self, y0, theta, return_sens=False, **kwargs):
         theta = at.cast(at.unbroadcast(at.as_tensor_variable(theta), 0), floatX)
         inputs = [y0, theta]
         for i, (input_val, itype) in enumerate(zip(inputs, self._itypes)):
-            if not input_val.type == itype:
+            if not input_val.type.in_same_class(itype):
                 raise ValueError(
                     f"Input {i} of type {input_val.type} does not have the expected type of {itype}"
                 )

pymc/tests/test_smc.py

@@ -140,20 +140,13 @@ def test_marginal_likelihood(self):
             with pm.Model() as model:
                 a = pm.Beta("a", alpha, beta)
                 y = pm.Bernoulli("y", a, observed=data)
-                trace = pm.sample_smc(2000, return_inferencedata=False)
-                marginals.append(trace.report.log_marginal_likelihood)
+                trace = pm.sample_smc(2000, chains=2, return_inferencedata=False)
+            # log_marignal_likelihood is found in the last value of each chain
+            lml = np.mean([chain[-1] for chain in trace.report.log_marginal_likelihood])
+            marginals.append(lml)
 
         # compare to the analytical result
-        assert (
-            np.abs(
-                np.exp(
-                    np.nanmean(np.array(marginals[1], dtype=float))
-                    - np.nanmean(np.array(marginals[0], dtype=float))
-                    - 4.0
-                )
-            )
-            <= 1
-        )
+        assert abs(np.exp(marginals[1] - marginals[0]) - 4.0) <= 1
 
     def test_start(self):
         with pm.Model() as model:

requirements-dev.txt

@@ -2,15 +2,15 @@
 # See that file for comments about the need/usage of each dependency.
 
 aeppl==0.0.18
-aesara==2.3.2
+aesara==2.3.6
 arviz>=0.11.4
 cachetools>=4.2.1
 cloudpickle
 fastprogress>=0.2.0
 h5py>=2.7
 ipython
 myst-nb
-numpy>=1.15.0,<1.22.0
+numpy>=1.15.0
 numpydoc>=0.9
 pandas
 pre-commit>=2.8.0

requirements.txt

@@ -1,10 +1,10 @@
 aeppl==0.0.18
-aesara==2.3.2
+aesara==2.3.6
 arviz>=0.11.4
 cachetools>=4.2.1
 cloudpickle
 fastprogress>=0.2.0
-numpy>=1.15.0,<1.22.0
+numpy>=1.15.0
 pandas>=0.24.0
 scipy>=1.4.1
 typing-extensions>=3.7.4