Add time values as sampler stats for NUTS

pymc3/step_methods/hmc/base_hmc.py

@@ -13,6 +13,7 @@
 #   limitations under the License.
 
 from collections import namedtuple
+import time
 
 import numpy as np
 import logging
@@ -132,6 +133,9 @@ def _hamiltonian_step(self, start, p0, step_size):
 
     def astep(self, q0):
         """Perform a single HMC iteration."""
+        perf_start = time.perf_counter_ns()
+        process_start = time.process_time_ns()
+
         p0 = self.potential.random()
         start = self.integrator.compute_state(q0, p0)
 
@@ -166,6 +170,9 @@ def astep(self, q0):
 
         hmc_step = self._hamiltonian_step(start, p0, step_size)
 
+        perf_end = time.perf_counter_ns()
+        process_end = time.process_time_ns()
+
         self.step_adapt.update(hmc_step.accept_stat, adapt_step)
         self.potential.update(hmc_step.end.q, hmc_step.end.q_grad, self.tune)
         if hmc_step.divergence_info:
@@ -191,7 +198,13 @@ def astep(self, q0):
         if not self.tune:
             self._samples_after_tune += 1
 
-        stats = {"tune": self.tune, "diverging": bool(hmc_step.divergence_info)}
+        stats = {
+            "tune": self.tune,
+            "diverging": bool(hmc_step.divergence_info),
+            "perf_counter_diff_ns": perf_end - perf_start,
+            "process_time_diff_ns": process_end - process_start,
+            "perf_counter_ns": perf_end,
+        }
 
         stats.update(hmc_step.stats)
         stats.update(self.step_adapt.stats())

pymc3/step_methods/hmc/hmc.py

@@ -48,6 +48,9 @@ class HamiltonianMC(BaseHMC):
         'path_length': np.float64,
         'accepted': np.bool,
         'model_logp': np.float64,
+        'process_time_diff_ns': np.int64,
+        'perf_counter_diff_ns': np.int64,
+        'perf_counter_ns': np.int64,
     }]
 
     def __init__(self, vars=None, path_length=2., max_steps=1024, **kwargs):

pymc3/step_methods/hmc/nuts.py

@@ -72,6 +72,13 @@ class NUTS(BaseHMC):
       samples, the step size is set to this value. This should converge
       during tuning.
     - `model_logp`: The model log-likelihood for this sample.
+    - `process_time_diff_ns`: The time it took to draw the sample, as defined
+      by the python standard library `time.process_time_ns`. This counts all
+      the CPU time, including worker processes in BLAS and OpenMP.
+    - `perf_counter_diff_ns`: The time it took to draw the sample, as defined
+      by the python standard library `time.perf_counter_ns` (wall time).
+    - `perf_counter_ns`: The value of the `time.perf_counter_ns` after drawing
+      the sample.
 
     References
     ----------
@@ -96,6 +103,9 @@ class NUTS(BaseHMC):
             "energy": np.float64,
             "max_energy_error": np.float64,
             "model_logp": np.float64,
+            "process_time_diff_ns": np.int64,
+            "perf_counter_diff_ns": np.int64,
+            "perf_counter_ns": np.int64,
         }
     ]