Use np.random's RandomState when seed is None

doc/source/whatsnew/v0.18.2.txt

@@ -56,6 +56,8 @@ API changes
 - Non-convertible dates in an excel date column will be returned without conversion and the column will be ``object`` dtype, rather than raising an exception  (:issue:`10001`)
 - Compat with ``np.round`` and timestamps (:issue:`12811`)
 - An ``UnsupportedFunctionCall`` error is now raised if numpy ufuncs like ``np.mean`` are called on groupby or resample objects (:issue:`12811`)
+- Calls to ``.sample()`` will respect random seed set via ``numpy.random.seed(n)`` (:issue:`13161`)
+
 
 .. _whatsnew_0182.api.tolist:
 

pandas/core/common.py

@@ -2062,7 +2062,7 @@ def _random_state(state=None):
     state : int, np.random.RandomState, None.
         If receives an int, passes to np.random.RandomState() as seed.
         If receives an np.random.RandomState object, just returns object.
-        If receives `None`, returns an np.random.RandomState object.
+        If receives `None`, returns np.random.
         If receives anything else, raises an informative ValueError.
         Default None.
 
@@ -2076,7 +2076,7 @@ def _random_state(state=None):
     elif isinstance(state, np.random.RandomState):
         return state
     elif state is None:
-        return np.random.RandomState()
+        return np.random
     else:
         raise ValueError("random_state must be an integer, a numpy "
                          "RandomState, or None")

pandas/tests/test_common.py

@@ -695,7 +695,7 @@ def test_random_state():
         com._random_state(state2).uniform(), npr.RandomState(10).uniform())
 
     # check with no arg random state
-    assert isinstance(com._random_state(), npr.RandomState)
+    assert com._random_state() is np.random
 
     # Error for floats or strings
     with tm.assertRaises(ValueError):

pandas/tests/test_generic.py

@@ -415,6 +415,14 @@ def test_sample(self):
                 o.sample(frac=0.7, random_state=np.random.RandomState(test)),
                 o.sample(frac=0.7, random_state=np.random.RandomState(test)))
 
+            os1, os2 = [], []
+            for _ in range(2):
+                np.random.seed(test)
+                os1.append(o.sample(n=4))
+                os2.append(o.sample(frac=0.7))
+            self._compare(*os1)
+            self._compare(*os2)
+
         # Check for error when random_state argument invalid.
         with tm.assertRaises(ValueError):
             o.sample(random_state='astring!')