def _infer_index_type(index):
""" Deduces native Numba type used to represent index Python object """
if isinstance(index, pd.RangeIndex):
# depending on actual index value unbox to diff types: none-index if it matches
# positions or to RangeIndexType in general case
if (index.start == 0 and index.step == 1 and index.name is None):
return types.none
else:
if index.name is None:
return RangeIndexType()
else:
return RangeIndexType(is_named=True)
# for unsupported pandas indexes we explicitly unbox to None
if isinstance(index, pd.DatetimeIndex):
return types.none
if index.dtype == np.dtype('O'):
# TO-DO: should we check that all elements are strings?
if len(index) > 0 and isinstance(index[0], str):
return string_array_type
else:
return types.none
numba_index_type = numpy_support.from_dtype(index.dtype)
return types.Array(numba_index_type, 1, 'C')
def _infer_index_type(index):
'''
Convertion input index type into Numba known type
need to return instance of the type class
'''
if isinstance(index, (types.NoneType, pd.RangeIndex, pd.DatetimeIndex)) or index is None or len(index) == 0:
return types.none
if index.dtype == np.dtype('O') and len(index) > 0:
first_val = index[0]
if isinstance(first_val, str):
return string_array_type
numba_index_type = numpy_support.from_dtype(index.dtype)
return types.Array(numba_index_type, 1, 'C')
