Merge pull request #6177 from wabu/selection-start-stop-fixes

ENH: select column/coordinates/multiple with start/stop/selection

Author: jreback

doc/source/release.rst

@@ -60,6 +60,7 @@ API Changes
   indexed. These will be excluded. This will make pandas conform more with pandas/numpy indexing of out-of-bounds
   values. A single indexer that is out-of-bounds and drops the dimensions of the object will still raise
   ``IndexError`` (:issue:`6296`)
+- ``select_as_multiple`` will always raise a ``KeyError``, when a key or the selector is not found (:issue:`6177`)
 
 Experimental Features
 ~~~~~~~~~~~~~~~~~~~~~
@@ -86,6 +87,9 @@ Bug Fixes
 - Bug in conversion of a string types to a DatetimeIndex with a specified frequency (:issue:`6273`, :issue:`6274`)
 - Bug in ``eval`` where type-promotion failed for large expressions (:issue:`6205`)
 - Bug in interpolate with inplace=True (:issue:`6281`)
+- ``HDFStore.remove`` now handles start and stop (:issue:`6177`)
+- ``HDFStore.select_as_multiple`` handles start and stop the same way as ``select`` (:issue:`6177`)
+- ``HDFStore.select_as_coordinates`` and ``select_column`` works where clauses that result in filters (:issue:`6177`)
 
 pandas 0.13.1
 -------------

pandas/io/pytables.py

@@ -724,8 +724,9 @@ def select_as_multiple(self, keys, where=None, selector=None, columns=None,
 
         Exceptions
         ----------
-        raise if any of the keys don't refer to tables or if they are not ALL
-        THE SAME DIMENSIONS
+        raises KeyError if keys or selector is not found or keys is empty
+        raises TypeError if keys is not a list or tuple
+        raises ValueError if the tables are not ALL THE SAME DIMENSIONS
         """
 
         # default to single select
@@ -748,12 +749,13 @@ def select_as_multiple(self, keys, where=None, selector=None, columns=None,
 
         # collect the tables
         tbls = [self.get_storer(k) for k in keys]
+        s = self.get_storer(selector)
 
         # validate rows
         nrows = None
-        for t, k in zip(tbls, keys):
+        for t, k in itertools.chain([(s,selector)], zip(tbls, keys)):
             if t is None:
-                raise TypeError("Invalid table [%s]" % k)
+                raise KeyError("Invalid table [%s]" % k)
             if not t.is_table:
                 raise TypeError(
                     "object [%s] is not a table, and cannot be used in all "
@@ -766,22 +768,17 @@ def select_as_multiple(self, keys, where=None, selector=None, columns=None,
                 raise ValueError(
                     "all tables must have exactly the same nrows!")
 
-        # select coordinates from the selector table
-        try:
-            c = self.select_as_coordinates(
-                selector, where, start=start, stop=stop)
-            nrows = len(c)
-        except Exception:
-            raise ValueError("invalid selector [%s]" % selector)
+        # axis is the concentation axes
+        axis = list(set([t.non_index_axes[0][0] for t in tbls]))[0]
 
         def func(_start, _stop):
-
-            # collect the returns objs
-            objs = [t.read(where=c[_start:_stop], columns=columns)
-                    for t in tbls]
-
-            # axis is the concentation axes
-            axis = list(set([t.non_index_axes[0][0] for t in tbls]))[0]
+            if where is not None:
+                c = s.read_coordinates(where=where, start=_start, stop=_stop, **kwargs)
+            else:
+                c = None
+                
+            objs = [t.read(where=c, start=_start, stop=_stop, 
+                           columns=columns, **kwargs) for t in tbls]
 
             # concat and return
             return concat(objs, axis=axis,
@@ -860,7 +857,7 @@ def remove(self, key, where=None, start=None, stop=None):
             raise KeyError('No object named %s in the file' % key)
 
         # remove the node
-        if where is None:
+        if where is None and start is None and stop is None:
             s.group._f_remove(recursive=True)
 
         # delete from the table
@@ -2139,11 +2136,9 @@ def write(self, **kwargs):
         raise NotImplementedError(
             "cannot write on an abstract storer: sublcasses should implement")
 
-    def delete(self, where=None, **kwargs):
-        """support fully deleting the node in its entirety (only) - where
-        specification must be None
-        """
-        if where is None:
+    def delete(self, where=None, start=None, stop=None, **kwargs):
+        """ support fully deleting the node in its entirety (only) - where specification must be None """
+        if where is None and start is None and stop is None:
             self._handle.removeNode(self.group, recursive=True)
             return None
 
@@ -3381,9 +3376,15 @@ def read_coordinates(self, where=None, start=None, stop=None, **kwargs):
         # create the selection
         self.selection = Selection(
             self, where=where, start=start, stop=stop, **kwargs)
-        return Index(self.selection.select_coords())
+        coords = self.selection.select_coords()
+        if self.selection.filter is not None:
+            for field, op, filt in self.selection.filter.format():
+                data = self.read_column(field, start=coords.min(), stop=coords.max()+1)
+                coords = coords[op(data.iloc[coords-coords.min()], filt).values]
 
-    def read_column(self, column, where=None, **kwargs):
+        return Index(coords)
+
+    def read_column(self, column, where=None, start=None, stop=None, **kwargs):
         """return a single column from the table, generally only indexables
         are interesting
         """
@@ -3411,7 +3412,7 @@ def read_column(self, column, where=None, **kwargs):
                 # column must be an indexable or a data column
                 c = getattr(self.table.cols, column)
                 a.set_info(self.info)
-                return Series(a.convert(c[:], nan_rep=self.nan_rep,
+                return Series(a.convert(c[start:stop], nan_rep=self.nan_rep,
                                         encoding=self.encoding).take_data())
 
         raise KeyError("column [%s] not found in the table" % column)
@@ -3712,12 +3713,19 @@ def write_data_chunk(self, indexes, mask, values):
         except Exception as detail:
             raise TypeError("tables cannot write this data -> %s" % detail)
 
-    def delete(self, where=None, **kwargs):
+    def delete(self, where=None, start=None, stop=None, **kwargs):
 
         # delete all rows (and return the nrows)
         if where is None or not len(where):
-            nrows = self.nrows
-            self._handle.removeNode(self.group, recursive=True)
+            if start is None and stop is None:
+                nrows = self.nrows
+                self._handle.removeNode(self.group, recursive=True)
+            else:
+                # pytables<3.0 would remove a single row with stop=None
+                if stop is None:
+                    stop = self.nrows
+                nrows = self.table.removeRows(start=start, stop=stop)
+                self.table.flush()
             return nrows
 
         # infer the data kind
@@ -3726,7 +3734,7 @@ def delete(self, where=None, **kwargs):
 
         # create the selection
         table = self.table
-        self.selection = Selection(self, where, **kwargs)
+        self.selection = Selection(self, where, start=start, stop=stop, **kwargs)
         values = self.selection.select_coords()
 
         # delete the rows in reverse order
@@ -4303,13 +4311,25 @@ def select_coords(self):
         """
         generate the selection
         """
-        if self.condition is None:
-            return np.arange(self.table.nrows)
+        start, stop = self.start, self.stop
+        nrows = self.table.nrows
+        if start is None:
+            start = 0
+        elif start < 0:
+            start += nrows
+        if self.stop is None:
+            stop = nrows
+        elif stop < 0:
+            stop += nrows
 
-        return self.table.table.getWhereList(self.condition.format(),
-                                             start=self.start, stop=self.stop,
-                                             sort=True)
+        if self.condition is not None:
+            return self.table.table.getWhereList(self.condition.format(),
+                                                 start=start, stop=stop,
+                                                 sort=True)
+        elif self.coordinates is not None:
+            return self.coordinates
 
+        return np.arange(start, stop)
 
 # utilities ###
 

pandas/io/tests/test_pytables.py

@@ -2195,6 +2195,69 @@ def test_remove_where(self):
             # self.assertRaises(ValueError, store.remove,
             #                  'wp2', [('column', ['A', 'D'])])
 
+    def test_remove_startstop(self):
+        # GH #4835 and #6177
+
+        with ensure_clean_store(self.path) as store:
+            
+            wp = tm.makePanel()
+
+            # start
+            store.put('wp1', wp, format='t')
+            n = store.remove('wp1', start=32)
+            #assert(n == 120-32)
+            result = store.select('wp1')
+            expected = wp.reindex(major_axis=wp.major_axis[:32//4])
+            assert_panel_equal(result, expected)
+
+            store.put('wp2', wp, format='t')
+            n = store.remove('wp2', start=-32)
+            #assert(n == 32)
+            result = store.select('wp2')
+            expected = wp.reindex(major_axis=wp.major_axis[:-32//4])
+            assert_panel_equal(result, expected)
+
+            # stop
+            store.put('wp3', wp, format='t')
+            n = store.remove('wp3', stop=32)
+            #assert(n == 32)
+            result = store.select('wp3')
+            expected = wp.reindex(major_axis=wp.major_axis[32//4:])
+            assert_panel_equal(result, expected)
+
+            store.put('wp4', wp, format='t')
+            n = store.remove('wp4', stop=-32)
+            #assert(n == 120-32)
+            result = store.select('wp4')
+            expected = wp.reindex(major_axis=wp.major_axis[-32//4:])
+            assert_panel_equal(result, expected)
+
+            # start n stop
+            store.put('wp5', wp, format='t')
+            n = store.remove('wp5', start=16, stop=-16)
+            #assert(n == 120-32)
+            result = store.select('wp5')
+            expected = wp.reindex(major_axis=wp.major_axis[:16//4]+wp.major_axis[-16//4:])
+            assert_panel_equal(result, expected)
+
+            store.put('wp6', wp, format='t')
+            n = store.remove('wp6', start=16, stop=16)
+            #assert(n == 0)
+            result = store.select('wp6')
+            expected = wp.reindex(major_axis=wp.major_axis)
+            assert_panel_equal(result, expected)
+            
+            # with where
+            date = wp.major_axis.take(np.arange(0,30,3))
+            crit = Term('major_axis=date')
+            store.put('wp7', wp, format='t')
+            n = store.remove('wp7', where=[crit], stop=80)
+            #assert(n == 28)
+            result = store.select('wp7')
+            expected = wp.reindex(major_axis=wp.major_axis-wp.major_axis[np.arange(0,20,3)])
+            assert_panel_equal(result, expected)
+            
+
     def test_remove_crit(self):
 
         with ensure_clean_store(self.path) as store:
@@ -3449,6 +3512,25 @@ def f():
             result = store.select_column('df3', 'string')
             tm.assert_almost_equal(result.values, df3['string'].values)
 
+            # start/stop
+            result = store.select_column('df3', 'string', start=2)
+            tm.assert_almost_equal(result.values, df3['string'].values[2:])
+
+            result = store.select_column('df3', 'string', start=-2)
+            tm.assert_almost_equal(result.values, df3['string'].values[-2:])
+
+            result = store.select_column('df3', 'string', stop=2)
+            tm.assert_almost_equal(result.values, df3['string'].values[:2])
+
+            result = store.select_column('df3', 'string', stop=-2)
+            tm.assert_almost_equal(result.values, df3['string'].values[:-2])
+
+            result = store.select_column('df3', 'string', start=2, stop=-2)
+            tm.assert_almost_equal(result.values, df3['string'].values[2:-2])
+
+            result = store.select_column('df3', 'string', start=-2, stop=2)
+            tm.assert_almost_equal(result.values, df3['string'].values[-2:2])
+
     def test_coordinates(self):
         df = tm.makeTimeDataFrame()
 
@@ -3519,6 +3601,12 @@ def test_coordinates(self):
             self.assertRaises(ValueError, store.select, 'df',where=np.arange(len(df)),start=5)
             self.assertRaises(ValueError, store.select, 'df',where=np.arange(len(df)),start=5,stop=10)
 
+            # selection with filter
+            selection = date_range('20000101',periods=500)
+            result = store.select('df', where='index in selection')
+            expected = df[df.index.isin(selection)]
+            tm.assert_frame_equal(result,expected)
+
             # list
             df = DataFrame(np.random.randn(10,2))
             store.append('df2',df)
@@ -3533,6 +3621,11 @@ def test_coordinates(self):
             expected = df.loc[where]
             tm.assert_frame_equal(result,expected)
 
+            # start/stop
+            result = store.select('df2', start=5, stop=10)
+            expected = df[5:10]
+            tm.assert_frame_equal(result,expected)
+
     def test_append_to_multiple(self):
         df1 = tm.makeTimeDataFrame()
         df2 = tm.makeTimeDataFrame().rename(columns=lambda x: "%s_2" % x)
@@ -3603,11 +3696,11 @@ def test_select_as_multiple(self):
                               None, where=['A>0', 'B>0'], selector='df1')
             self.assertRaises(Exception, store.select_as_multiple,
                               [None], where=['A>0', 'B>0'], selector='df1')
-            self.assertRaises(TypeError, store.select_as_multiple,
+            self.assertRaises(KeyError, store.select_as_multiple,
                               ['df1','df3'], where=['A>0', 'B>0'], selector='df1')
             self.assertRaises(KeyError, store.select_as_multiple,
                               ['df3'], where=['A>0', 'B>0'], selector='df1')
-            self.assertRaises(ValueError, store.select_as_multiple,
+            self.assertRaises(KeyError, store.select_as_multiple,
                               ['df1','df2'], where=['A>0', 'B>0'], selector='df4')
 
             # default select