set a different default executor for Windows and Unix systems

A 'concurrent.futures.ProcessPoolExecutor' doesn't work on Windows
so we use 'distributed.Client'.

adaptive/runner.py

@@ -1,10 +1,11 @@
 # -*- coding: utf-8 -*-
 import asyncio
+import concurrent.futures as concurrent
 import functools
 import inspect
-import concurrent.futures as concurrent
-import warnings
+import os
 import time
+import warnings
 
 from .notebook_integration import live_plot, live_info, in_ipynb
 
@@ -27,6 +28,23 @@
     pass
 
 
+if os.name == 'nt':
+    if with_distributed:
+        _default_executor = distributed.Client
+        _default_executor_kwargs = {'address': distributed.LocalCluster()}
+    else:
+        warnings.warn('The default executor on Windows (`distributed.Client`) '
+                      'for adaptive.Runner cannot be used because `distributed` is not '
+                      'installed. Consider installing `distributed`. '
+                      'Specify `executor` when using the `Runner`. Note: a '
+                      '`concurrent.futures.ProcessPoolExecutor` does not work on '
+                      'Windows.')
+else:
+    _default_executor = concurrent.ProcessPoolExecutor
+    _default_executor_kwargs = {}
+
+
+
 class BaseRunner:
     """Base class for runners that use concurrent.futures.Executors.
 
@@ -86,9 +104,12 @@ class BlockingRunner(BaseRunner):
         The end condition for the calculation. This function must take
         the learner as its sole argument, and return True when we should
         stop requesting more points.
-    executor : concurrent.futures.Executor, or ipyparallel.Client, optional
+    executor : concurrent.futures.Executor, distributed.Client,
+               or ipyparallel.Client, optional
         The executor in which to evaluate the function to be learned.
-        If not provided, a new ProcessPoolExecutor is used.
+        If not provided, a new `ProcessPoolExecutor` is used on Unix systems
+        while on Windows a `distributed.Client` is used if `distributed` is
+        installed.
     ntasks : int, optional
         The number of concurrent function evaluations. Defaults to the number
         of cores available in 'executor'.
@@ -183,9 +204,12 @@ class AsyncRunner(BaseRunner):
         the learner as its sole argument, and return True when we should
         stop requesting more points. If not provided, the runner will run
         forever, or until 'self.task.cancel()' is called.
-    executor : concurrent.futures.Executor, or ipyparallel.Client, optional
+    executor : concurrent.futures.Executor, distributed.Client,
+               or ipyparallel.Client, optional
         The executor in which to evaluate the function to be learned.
-        If not provided, a new ProcessPoolExecutor is used.
+        If not provided, a new `ProcessPoolExecutor` is used on Unix systems
+        while on Windows a `distributed.Client` is used if `distributed` is
+        installed.
     ntasks : int, optional
         The number of concurrent function evaluations. Defaults to the number
         of cores available in 'executor'.
@@ -440,8 +464,9 @@ def shutdown(self, wait=True):
 
 def _ensure_executor(executor):
     if executor is None:
-        return concurrent.ProcessPoolExecutor()
-    elif isinstance(executor, concurrent.Executor):
+        executor = _default_executor(**_default_executor_kwargs)
+
+    if isinstance(executor, concurrent.Executor):
         return executor
     elif with_ipyparallel and isinstance(executor, ipyparallel.Client):
         return executor.executor()