Doc pdist's keyword arguments V and VI

Missed documenting `V` and `VI` when adding support for the `seuclidean`
and `mahalanobis` metrics to `pdist` respectively. So this adds
documentation for these parameters along with a brief explanation of the
default values they can take on if not specified.

dask_distance/__init__.py

@@ -132,6 +132,9 @@ def pdist(X, metric="euclidean", **kwargs):
 
     Keyword Args:
         p:          p-norm for minkowski only (default: 2)
+        V:          1-D array of variances (default: estimated from X)
+        VI:         Inverse of the covariance matrix
+                    (default: estimated from X)
 
     Returns:
         array:      condensed distance between each pair