Merge pull request #711 from ioam/dataset_dynamic_groupby

Added support for dynamic groupby on all data interfaces

holoviews/core/data/__init__.py

@@ -48,7 +48,7 @@
 
 from ..dimension import Dimension
 from ..element import Element
-from ..spaces import HoloMap
+from ..spaces import HoloMap, DynamicMap
 from .. import util
 
 
@@ -369,26 +369,42 @@ def aggregate(self, dimensions=None, function=None, spreadfn=None, **kwargs):
 
 
 
-    def groupby(self, dimensions=[], container_type=HoloMap, group_type=None, **kwargs):
-        """
-        Return the results of a groupby operation over the specified
+    def groupby(self, dimensions=[], container_type=HoloMap, group_type=None,
+                dynamic=False, **kwargs):
+        """Return the results of a groupby operation over the specified
         dimensions as an object of type container_type (expected to be
         dictionary-like).
 
         Keys vary over the columns (dimensions) and the corresponding
-        values are collections of group_type (e.g list, tuple)
+        values are collections of group_type (e.g an Element, list, tuple)
         constructed with kwargs (if supplied).
+
+        If dynamic is requested container_type is automatically set to
+        a DynamicMap, allowing dynamic exploration of large
+        datasets. If the data does not represent a full cartesian grid
+        of the requested dimensions some Elements will be empty.
         """
         if not isinstance(dimensions, list): dimensions = [dimensions]
         if not len(dimensions): dimensions = self.dimensions('key', True)
         if group_type is None: group_type = type(self)
 
-        dimensions = [self.get_dimension(d, strict=True).name for d in dimensions]
-        invalid_dims = list(set(dimensions) - set(self.dimensions('key', True)))
-        if invalid_dims:
-            raise Exception('Following dimensions could not be found:\n%s.'
-                            % invalid_dims)
-        return self.interface.groupby(self, dimensions, container_type,
+        dimensions = [self.get_dimension(d, strict=True) for d in dimensions]
+        dim_names = [d.name for d in dimensions]
+
+        if dynamic:
+            group_dims = [d.name for d in self.kdims if d not in dimensions]
+            def load_subset(*args):
+                constraint = dict(zip(dim_names, args))
+                group = self.select(**constraint)
+                if np.isscalar(group):
+                    return group_type(([group],), group=self.group,
+                                      label=self.label, vdims=self.vdims)
+                return group_type(group.reindex(group_dims))
+            dynamic_dims = [d(values=list(self.interface.values(self, d.name, False)))
+                            for d in dimensions]
+            return DynamicMap(load_subset, kdims=dynamic_dims)
+
+        return self.interface.groupby(self, dim_names, container_type,
                                       group_type, **kwargs)
 
     def __len__(self):

holoviews/core/data/grid.py

@@ -291,7 +291,9 @@ def reindex(cls, dataset, kdims, vdims):
                 vdata = data[vdim.name]
                 if dropped_axes:
                     vdata = vdata.squeeze(axis=dropped_axes)
-                data[vdim.name] = np.transpose(vdata, axes)
+                if len(axes) > 1:
+                    vdata = np.transpose(vdata, axes)
+                data[vdim.name] = vdata
         return data
 
 

holoviews/core/data/iris.py

@@ -148,6 +148,14 @@ def reindex(cls, dataset, kdims=None, vdims=None):
         does not need to be reindexed, the Element can simply
         reorder its key dimensions.
         """
+        if kdims and len(kdims) != dataset.ndims:
+            drop_dims = [kd for kd in dataset.kdims if kd not in kdims]
+            constraints = {}
+            for d in drop_dims:
+                vals = cls.values(dataset, d, False)
+                if len(vals):
+                    constraints[d.name] = vals[0]
+            return dataset.data.extract(iris.Constraint(**constraints))
         return dataset.data
 
 
@@ -165,15 +173,6 @@ def groupby(cls, dataset, dims, container_type=HoloMap, group_type=None, **kwarg
         constraints = [d.name for d in dims]
         slice_dims = [d for d in dataset.kdims if d not in dims]
 
-        if dynamic:
-            def load_subset(*args):
-                constraint = iris.Constraint(**dict(zip(constraints, args)))
-                return dataset.clone(dataset.data.extract(constraint),
-                                      new_type=group_type,
-                                      **dict(kwargs, kdims=slice_dims))
-            dynamic_dims = [d(values=list(cls.values(dataset, d, False))) for d in dims]
-            return DynamicMap(load_subset, kdims=dynamic_dims)
-
         unique_coords = product(*[cls.values(dataset, d, expanded=False)
                                   for d in dims])
         data = []

tests/testdataset.py

@@ -280,6 +280,12 @@ def test_dataset_groupby(self):
                           kdims=['Gender'])
         self.assertEqual(self.table.groupby(['Gender']), grouped)
 
+    def test_dataset_groupby_dynamic(self):
+        grouped_dataset = self.table.groupby('Gender', dynamic=True)
+        self.assertEqual(grouped_dataset['M'],
+                         self.table.select(Gender='M').reindex(['Age']))
+        self.assertEqual(grouped_dataset['F'],
+                         self.table.select(Gender='F').reindex(['Age']))
 
     def test_dataset_add_dimensions_value_ht(self):
         table = self.dataset_ht.add_dimension('z', 1, 0)
@@ -491,6 +497,15 @@ def test_dataset_sort_vdim_hm(self):
     def test_dataset_groupby(self):
         self.assertEqual(self.dataset_hm.groupby('x').keys(), list(self.xs))
 
+    def test_dataset_groupby_dynamic(self):
+        array = np.random.rand(11, 11)
+        dataset = Dataset({'x':self.xs, 'y':self.y_ints, 'z': array},
+                          kdims=['x', 'y'], vdims=['z'])
+        grouped = dataset.groupby('x', dynamic=True)
+        first = Dataset({'y': self.y_ints, 'z': array[:, 0]},
+                        kdims=['y'], vdims=['z'])
+        self.assertEqual(grouped[0], first)
+
 
 
 class IrisDatasetTest(GridDatasetTest):
@@ -525,7 +540,6 @@ def test_dataset_sample_hm(self):
         pass
 
 
-
 class XArrayDatasetTest(GridDatasetTest):
     """
     Tests for Iris interface