ENH: raise AmbiguousIndexError when location not found in iteger-indexed Series. Start of #328 API changes

Author: wesm

pandas/core/common.py

@@ -27,6 +27,9 @@
 class PandasError(Exception):
     pass
 
+class AmbiguousIndexError(PandasError, KeyError):
+    pass
+
 def isnull(obj):
     '''
     Replacement for numpy.isnan / -numpy.isfinite which is suitable

pandas/core/frame.py

@@ -1368,7 +1368,7 @@ def _sanitize_column(self, value):
         else:
             value = np.repeat(value, len(self.index))
 
-        return value
+        return np.asarray(value)
 
     def pop(self, item):
         """
@@ -1729,7 +1729,9 @@ def reset_index(self):
 
                 # to ndarray and maybe infer different dtype
                 level_values = lev.values
-                level_values = lib.maybe_convert_objects(level_values)
+                if level_values.dtype == np.object_:
+                    level_values = lib.maybe_convert_objects(level_values)
+
                 new_obj.insert(0, col_name, level_values.take(lab))
         else:
             name = self.index.name

pandas/core/index.py

@@ -66,6 +66,9 @@ def __new__(cls, data, dtype=None, copy=False, name=None):
             # other iterable of some kind
             subarr = _asarray_tuplesafe(data, dtype=object)
 
+        if lib.is_integer_array(subarr) and dtype is None:
+            return Int64Index(subarr.astype('i8'), name=name)
+
         subarr = subarr.view(cls)
         subarr.name = name
         return subarr
@@ -74,7 +77,8 @@ def __array_finalize__(self, obj):
         self.name = getattr(obj, 'name', None)
 
     def astype(self, dtype):
-        return Index(self.values.astype(dtype), name=self.name)
+        return Index(self.values.astype(dtype), name=self.name,
+                     dtype=dtype)
 
     @property
     def dtype(self):
@@ -857,13 +861,14 @@ def __new__(cls, data, dtype=None, copy=False, name=None):
         subarr.name = name
         return subarr
 
+    @property
+    def inferred_type(self):
+        return 'integer'
+
     @property
     def _constructor(self):
         return Int64Index
 
-    def astype(self, dtype):
-        return Index(self.values.astype(dtype))
-
     @property
     def dtype(self):
         return np.dtype('int64')

pandas/core/series.py

@@ -15,7 +15,8 @@
 
 from pandas.core.common import (isnull, notnull, _is_bool_indexer,
                                 _default_index, _maybe_upcast,
-                                _asarray_tuplesafe)
+                                _asarray_tuplesafe,
+                                AmbiguousIndexError)
 from pandas.core.daterange import DateRange
 from pandas.core.index import Index, MultiIndex, _ensure_index
 from pandas.core.indexing import _SeriesIndexer, _maybe_droplevels
@@ -137,6 +138,8 @@ class Series(np.ndarray, generic.PandasObject):
 
     _AXIS_NAMES = dict((v, k) for k, v in _AXIS_NUMBERS.iteritems())
 
+    __slots__ = ['_index', 'name']
+
     def __new__(cls, data=None, index=None, dtype=None, name=None,
                 copy=False):
         if data is None:
@@ -208,20 +211,22 @@ def __hash__(self):
         raise TypeError('unhashable type')
 
     _index = None
-    def _get_index(self):
-        return self._index
+    index = lib.SeriesIndex()
+
+    # def _get_index(self):
+    #     return self._index
 
-    def _set_index(self, index):
-        if not isinstance(index, _INDEX_TYPES):
-            raise TypeError("Expected index to be in %s; was %s."
-                            % (_INDEX_TYPES, type(index)))
+    # def _set_index(self, index):
+    #     if not isinstance(index, _INDEX_TYPES):
+    #         raise TypeError("Expected index to be in %s; was %s."
+    #                         % (_INDEX_TYPES, type(index)))
 
-        if len(self) != len(index):
-            raise AssertionError('Lengths of index and values did not match!')
+    #     if len(self) != len(index):
+    #         raise AssertionError('Lengths of index and values did not match!')
 
-        self._index = _ensure_index(index)
+    #     self._index = _ensure_index(index)
 
-    index = property(fget=_get_index, fset=_set_index)
+    # index = property(fget=_get_index, fset=_set_index)
 
     def __array_finalize__(self, obj):
         """
@@ -264,22 +269,27 @@ def ix(self):
         return self._ix
 
     def __getitem__(self, key):
+        index = self.index
+
         # Label-based
         try:
-            return self.index._engine.get_value(self, key)
+            return index._engine.get_value(self, key)
         except KeyError, e1:
-            if isinstance(self.index, MultiIndex):
+            if isinstance(index, MultiIndex):
                 values = self.values
                 try:
-                    loc = self.index.get_loc(key)
+                    loc = index.get_loc(key)
                     # TODO: what if a level contains tuples??
-                    new_index = self.index[loc]
+                    new_index = index[loc]
                     new_index = _maybe_droplevels(new_index, key)
                     return Series(values[loc], index=new_index,
                                   name=self.name)
                 except KeyError:
                     pass
 
+            if index.inferred_type == 'integer':
+                raise AmbiguousIndexError(key)
+
             try:
                 return _gin.get_value_at(self, key)
             except IndexError:

pandas/sparse/tests/test_sparse.py

@@ -286,9 +286,10 @@ def _check_getitem(sp, dense):
                 # j = np.float64(i)
                 # assert_almost_equal(sp[j], dense[j])
 
+            # API change 1/6/2012
             # negative getitem works
-            for i in xrange(len(dense)):
-                assert_almost_equal(sp[-i], dense[-i])
+            # for i in xrange(len(dense)):
+            #     assert_almost_equal(sp[-i], dense[-i])
 
         _check_getitem(self.bseries, self.bseries.to_dense())
         _check_getitem(self.btseries, self.btseries.to_dense())
@@ -563,7 +564,8 @@ def test_valid(self):
                           fill_value=0)
 
         sp_valid = sp.valid()
-        assert_almost_equal(sp_valid, sp.to_dense().valid())
+        assert_almost_equal(sp_valid.values,
+                            sp.to_dense().valid().values)
         self.assert_(sp_valid.index.equals(sp.to_dense().valid().index))
         self.assertEquals(len(sp_valid.sp_values), 2)
 

pandas/src/properties.pyx

@@ -41,3 +41,20 @@ cdef class AxisProperty(object):
 
     def __set__(self, obj, value):
         obj._set_axis(self.axis, value)
+
+cdef class SeriesIndex(object):
+    cdef:
+        Py_ssize_t axis
+        object _check_type
+
+    def __init__(self):
+        from pandas.core.index import _ensure_index
+        self._check_type = _ensure_index
+
+    def __get__(self, obj, type):
+        return obj._index
+
+    def __set__(self, obj, value):
+        if len(obj) != len(value):
+            raise AssertionError('Index length did not match values')
+        obj._index = self._check_type(value)

pandas/tests/test_index.py

@@ -404,6 +404,15 @@ def test_constructor_corner(self):
         arr = np.array([1, '2', 3, '4'], dtype=object)
         self.assertRaises(TypeError, Int64Index, arr)
 
+    def test_coerce_list(self):
+        # coerce things
+        arr = Index([1, 2, 3, 4])
+        self.assert_(type(arr) == Int64Index)
+
+        # but not if explicit dtype passed
+        arr = Index([1, 2, 3, 4], dtype=object)
+        self.assert_(type(arr) == Index)
+
     def test_dtype(self):
         self.assert_(self.index.dtype == np.int64)
 

pandas/tests/test_series.py

@@ -262,11 +262,11 @@ def test_fromDict(self):
     def test_setindex(self):
         # wrong type
         series = self.series.copy()
-        self.assertRaises(TypeError, series._set_index, None)
+        self.assertRaises(TypeError, setattr, series, 'index', None)
 
         # wrong length
         series = self.series.copy()
-        self.assertRaises(AssertionError, series._set_index,
+        self.assertRaises(AssertionError, setattr, series, 'index',
                           np.arange(len(series) - 1))
 
         # works
@@ -400,6 +400,11 @@ def test_getitem_box_float64(self):
         value = self.ts[5]
         self.assert_(isinstance(value, np.float64))
 
+    def test_getitem_ambiguous_keyerror(self):
+        s = Series(range(10), index=range(0, 20, 2))
+        self.assertRaises(KeyError, s.__getitem__, 1)
+        self.assertRaises(KeyError, s.ix.__getitem__, 1)
+
     def test_slice(self):
         numSlice = self.series[10:20]
         numSliceEnd = self.series[-10:]

pandas/tests/test_tseries.py

@@ -27,31 +27,31 @@ def test_merge_indexer(self):
         old = Index([1, 5, 10])
         new = Index(range(12))
 
-        filler = lib.merge_indexer_object(new, old.indexMap)
+        filler = lib.merge_indexer_int64(new, old.indexMap)
 
         expect_filler = [-1, 0, -1, -1, -1, 1, -1, -1, -1, -1, 2, -1]
         self.assert_(np.array_equal(filler, expect_filler))
 
         # corner case
         old = Index([1, 4])
         new = Index(range(5, 10))
-        filler = lib.merge_indexer_object(new, old.indexMap)
+        filler = lib.merge_indexer_int64(new, old.indexMap)
         expect_filler = [-1, -1, -1, -1, -1]
         self.assert_(np.array_equal(filler, expect_filler))
 
     def test_backfill(self):
         old = Index([1, 5, 10])
         new = Index(range(12))
 
-        filler = lib.backfill_object(old, new, old.indexMap, new.indexMap)
+        filler = lib.backfill_int64(old, new, old.indexMap, new.indexMap)
 
         expect_filler = [0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 2, -1]
         self.assert_(np.array_equal(filler, expect_filler))
 
         # corner case
         old = Index([1, 4])
         new = Index(range(5, 10))
-        filler = lib.backfill_object(old, new, old.indexMap, new.indexMap)
+        filler = lib.backfill_int64(old, new, old.indexMap, new.indexMap)
 
         expect_filler = [-1, -1, -1, -1, -1]
         self.assert_(np.array_equal(filler, expect_filler))
@@ -60,15 +60,15 @@ def test_pad(self):
         old = Index([1, 5, 10])
         new = Index(range(12))
 
-        filler = lib.pad_object(old, new, old.indexMap, new.indexMap)
+        filler = lib.pad_int64(old, new, old.indexMap, new.indexMap)
 
         expect_filler = [-1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 2, 2]
         self.assert_(np.array_equal(filler, expect_filler))
 
         # corner case
         old = Index([5, 10])
         new = Index(range(5))
-        filler = lib.pad_object(old, new, old.indexMap, new.indexMap)
+        filler = lib.pad_int64(old, new, old.indexMap, new.indexMap)
         expect_filler = [-1, -1, -1, -1, -1]
         self.assert_(np.array_equal(filler, expect_filler))