Merge branch 'master' into docs/add-citation-from-2012.12246

src/pyhf/tensor/pytorch_backend.py

@@ -319,7 +319,8 @@ def einsum(self, subscripts, *operands):
         return torch.einsum(subscripts, operands)
 
     def poisson_logpdf(self, n, lam):
-        return torch.distributions.Poisson(lam).log_prob(n)
+        # validate_args=True disallows continuous approximation
+        return torch.distributions.Poisson(lam, validate_args=False).log_prob(n)
 
     def poisson(self, n, lam):
         r"""
@@ -347,9 +348,16 @@ def poisson(self, n, lam):
         Returns:
             PyTorch FloatTensor: Value of the continuous approximation to Poisson(n|lam)
         """
-        return torch.exp(torch.distributions.Poisson(lam).log_prob(n))
+        # validate_args=True disallows continuous approximation
+        return torch.exp(
+            torch.distributions.Poisson(lam, validate_args=False).log_prob(n)
+        )
 
     def normal_logpdf(self, x, mu, sigma):
+        x = self.astensor(x)
+        mu = self.astensor(mu)
+        sigma = self.astensor(sigma)
+
         normal = torch.distributions.Normal(mu, sigma)
         return normal.log_prob(x)
 
@@ -379,6 +387,10 @@ def normal(self, x, mu, sigma):
         Returns:
             PyTorch FloatTensor: Value of Normal(x|mu, sigma)
         """
+        x = self.astensor(x)
+        mu = self.astensor(mu)
+        sigma = self.astensor(sigma)
+
         normal = torch.distributions.Normal(mu, sigma)
         return self.exp(normal.log_prob(x))
 
@@ -433,7 +445,8 @@ def poisson_dist(self, rate):
             PyTorch Poisson distribution: The Poisson distribution class
 
         """
-        return torch.distributions.Poisson(rate)
+        # validate_args=True disallows continuous approximation
+        return torch.distributions.Poisson(rate, validate_args=False)
 
     def normal_dist(self, mu, sigma):
         r"""

tests/test_tensor.py

@@ -247,6 +247,8 @@ def test_shape(backend):
         )
 
 
+@pytest.mark.skip_pytorch
+@pytest.mark.skip_pytorch64
 def test_pdf_calculations(backend):
     tb = pyhf.tensorlib
     assert tb.tolist(tb.normal_cdf(tb.astensor([0.8]))) == pytest.approx(
@@ -286,7 +288,53 @@ def test_pdf_calculations(backend):
 
     # Ensure continuous approximation is valid
     assert tb.tolist(
-        tb.poisson(tb.astensor([0.5, 1.1, 1.5]), tb.astensor(1.0))
+        tb.poisson(n=tb.astensor([0.5, 1.1, 1.5]), lam=tb.astensor(1.0))
+    ) == pytest.approx([0.4151074974205947, 0.3515379040027489, 0.2767383316137298])
+
+
+# validate_args in torch.distributions raises ValueError not nan
+@pytest.mark.only_pytorch
+@pytest.mark.only_pytorch64
+def test_pdf_calculations_pytorch(backend):
+    tb = pyhf.tensorlib
+
+    values = tb.astensor([0, 0, 1, 1])
+    mus = tb.astensor([0, 1, 0, 1])
+    sigmas = tb.astensor([0, 0, 0, 0])
+    for x, mu, sigma in zip(values, mus, sigmas):
+        with pytest.raises(ValueError):
+            _ = tb.normal_logpdf(x, mu, sigma)
+    assert tb.tolist(
+        tb.normal_logpdf(
+            tb.astensor([0, 0, 1, 1]),
+            tb.astensor([0, 1, 0, 1]),
+            tb.astensor([1, 1, 1, 1]),
+        )
+    ) == pytest.approx(
+        [
+            -0.91893853,
+            -1.41893853,
+            -1.41893853,
+            -0.91893853,
+        ],
+    )
+
+    # poisson(lambda=0) is not defined, should return NaN
+    assert tb.tolist(
+        tb.poisson(tb.astensor([0, 0, 1, 1]), tb.astensor([0, 1, 0, 1]))
+    ) == pytest.approx(
+        [np.nan, 0.3678794503211975, 0.0, 0.3678794503211975], nan_ok=True
+    )
+    assert tb.tolist(
+        tb.poisson_logpdf(tb.astensor([0, 0, 1, 1]), tb.astensor([0, 1, 0, 1]))
+    ) == pytest.approx(
+        np.log([np.nan, 0.3678794503211975, 0.0, 0.3678794503211975]).tolist(),
+        nan_ok=True,
+    )
+
+    # Ensure continuous approximation is valid
+    assert tb.tolist(
+        tb.poisson(n=tb.astensor([0.5, 1.1, 1.5]), lam=tb.astensor(1.0))
     ) == pytest.approx([0.4151074974205947, 0.3515379040027489, 0.2767383316137298])