- Fix regression caused by #4211 (#4285)

* - Fix regression caused by #4211

* - Add test to make sure jitter is being applied to chains starting points by default

* - Import appropriate empty context for python < 3.7

* - Apply black formatting

* - Change the second check_start_vals to explicitly run on the newly assigned start variable.

* - Improve test documentation and add a new condition

* Use monkeypatch for more robust test

* - Black formatting, once again...

pymc3/sampling.py

@@ -416,15 +416,15 @@ def sample(
     """
     model = modelcontext(model)
     if start is None:
-        start = model.test_point
+        check_start_vals(model.test_point, model)
     else:
         if isinstance(start, dict):
             update_start_vals(start, model.test_point, model)
         else:
             for chain_start_vals in start:
                 update_start_vals(chain_start_vals, model.test_point, model)
+        check_start_vals(start, model)
 
-    check_start_vals(start, model)
     if cores is None:
         cores = min(4, _cpu_count())
 
@@ -492,9 +492,9 @@ def sample(
                 progressbar=progressbar,
                 **kwargs,
             )
-            check_start_vals(start_, model)
             if start is None:
                 start = start_
+                check_start_vals(start, model)
         except (AttributeError, NotImplementedError, tg.NullTypeGradError):
             # gradient computation failed
             _log.info("Initializing NUTS failed. " "Falling back to elementwise auto-assignment.")

pymc3/tests/test_sampling.py

@@ -12,6 +12,7 @@
 #   See the License for the specific language governing permissions and
 #   limitations under the License.
 
+from contextlib import ExitStack as does_not_raise
 from itertools import combinations
 from typing import Tuple
 import numpy as np
@@ -25,7 +26,7 @@
 import theano
 from pymc3.tests.models import simple_init
 from pymc3.tests.helpers import SeededTest
-from pymc3.exceptions import IncorrectArgumentsError
+from pymc3.exceptions import IncorrectArgumentsError, SamplingError
 from scipy import stats
 import pytest
 
@@ -785,6 +786,35 @@ def test_exec_nuts_init(method):
         assert "a" in start[0] and "b_log__" in start[0]
 
 
+@pytest.mark.parametrize(
+    "init, start, expectation",
+    [
+        ("auto", None, pytest.raises(SamplingError)),
+        ("jitter+adapt_diag", None, pytest.raises(SamplingError)),
+        ("auto", {"x": 0}, does_not_raise()),
+        ("jitter+adapt_diag", {"x": 0}, does_not_raise()),
+        ("adapt_diag", None, does_not_raise()),
+    ],
+)
+def test_default_sample_nuts_jitter(init, start, expectation, monkeypatch):
+    # This test tries to check whether the starting points returned by init_nuts are actually
+    # being used when pm.sample() is called without specifying an explicit start point (see
+    # https://github.com/pymc-devs/pymc3/pull/4285).
+    def _mocked_init_nuts(*args, **kwargs):
+        if init == "adapt_diag":
+            start_ = [{"x": np.array(0.79788456)}]
+        else:
+            start_ = [{"x": np.array(-0.04949886)}]
+        _, step = pm.init_nuts(*args, **kwargs)
+        return start_, step
+
+    monkeypatch.setattr("pymc3.sampling.init_nuts", _mocked_init_nuts)
+    with pm.Model() as m:
+        x = pm.HalfNormal("x", transform=None)
+        with expectation:
+            pm.sample(tune=1, draws=0, chains=1, init=init, start=start)
+
+
 @pytest.fixture(scope="class")
 def point_list_arg_bug_fixture() -> Tuple[pm.Model, pm.backends.base.MultiTrace]:
     with pm.Model() as pmodel: