MAINT: implement median

dask/array/__init__.py

@@ -40,7 +40,7 @@
                          nanmax, nanargmin, nanargmax,
                          cumsum, cumprod,
                          topk, argtopk)
-from .percentile import percentile
+from .percentile import percentile, median
 with ignoring(ImportError):
     from .reductions import nanprod, nancumprod, nancumsum
 with ignoring(ImportError):

dask/array/core.py

@@ -1747,6 +1747,10 @@ def mean(self, axis=None, dtype=None, keepdims=False, split_every=None,
         return mean(self, axis=axis, dtype=dtype, keepdims=keepdims,
                     split_every=split_every, out=out)
 
+    def median(self):
+        from .percentile import median
+        return median(self)
+
     @derived_from(np.ndarray)
     def std(self, axis=None, dtype=None, keepdims=False, ddof=0,
             split_every=None, out=None):

dask/array/percentile.py

@@ -61,6 +61,14 @@ def percentile(a, q, interpolation='linear'):
     return Array(dsk, name2, chunks=((len(q),),), dtype=dtype)
 
 
+def median(x, interpolation='linear'):
+    """ Approximate the median function
+
+    See :func:`percentile` for more information.
+    """
+    return percentile(x, 50, interpolation=interpolation)
+
+
 def merge_percentiles(finalq, qs, vals, interpolation='lower', Ns=None):
     """ Combine several percentile calculations of different data.
 

dask/array/tests/test_percentiles.py

@@ -28,6 +28,17 @@ def test_percentile():
               np.array(['a', 'd', 'e'], dtype=x.dtype))
 
 
+def test_median():
+    rng = np.random.RandomState(42)
+    n = 1000
+    x = rng.rand(n)
+    d = da.from_array(x, chunks=n // 10)
+
+    assert_eq(d.median(), da.median(d))
+    assert_eq(da.percentile(d, 50), da.median(d))
+    assert np.abs(da.median(d).compute() - np.median(x)) < 0.07
+
+
 @pytest.mark.skip
 def test_percentile_with_categoricals():
     try: