def fit(localization, proxy_obj, arguments): dvar = call_utilities.parse_varargs_and_kwargs( localization, arguments, ['domain', 'rcond', 'full', 'w', 'window'], { 'domain': IterableDataStructure, 'rcond': RealNumber, 'full': bool, 'w': IterableDataStructure, 'window': IterableDataStructure, }, 'fit', 3) if isinstance(dvar, StypyTypeError): return dvar ret = call_utilities.create_numpy_array_n_dimensions( call_utilities.get_inner_type(localization, arguments[0]), call_utilities.get_dimensions(localization, arguments[0])) ret = numpy.polynomial.Chebyshev(ret.get_wrapped_type()) if 'full' in dvar.keys(): tup = wrap_contained_type(tuple()) ld = wrap_contained_type(list()) ld.set_contained_type( call_utilities.get_inner_type(localization, arguments[0])) un = UnionType.add(ret, ld) tup.set_contained_type(un) return tup return ret
def __getitem__(localization, proxy_obj, arguments): r = TypeWrapper.get_wrapper_of(proxy_obj.__self__) index_selector = arguments[0] if isinstance(index_selector, tuple): index_selector = index_selector[0] dims = call_utilities.get_dimensions(localization, r) if dims > 1: if call_utilities.is_iterable(arguments[0]): if call_utilities.is_iterable(arguments[0].get_contained_type()): return call_utilities.create_numpy_array_n_dimensions( call_utilities.cast_to_numpy_type(call_utilities.get_inner_type(localization, r)), dims - 1) contained = call_utilities.cast_to_numpy_type(call_utilities.get_inner_type(localization, r)) for i in range(dims - 1): contained = UnionType.add(contained, call_utilities.create_numpy_array_n_dimensions(r.get_contained_type(), i + 1)) else: contained = r.get_contained_type() if isinstance(index_selector, TypeWrapper): if isinstance(index_selector.wrapped_type, slice) or ( call_utilities.is_iterable(index_selector) and not isinstance(index_selector.wrapped_type, tuple)): l = call_utilities.create_numpy_array(contained) return l return contained # proxy_obj.__self__.dtype.type()
def dot(localization, proxy_obj, arguments): dvar = call_utilities.parse_varargs_and_kwargs( localization, arguments, ['out'], { 'out': numpy.ndarray, }, 'dot') if isinstance(dvar, StypyTypeError): return dvar if Number == type(arguments[0]) and Number == type(arguments[1]): return call_utilities.cast_to_greater_numpy_type( arguments[0], arguments[1]) if Number == type(arguments[0]) and call_utilities.is_iterable( arguments[1]): c_t = get_contained_elements_type(localization, arguments[1]) if not 'out' in dvar.keys(): return call_utilities.create_numpy_array(c_t) else: set_contained_elements_type(localization, dvar['out'], c_t) return dvar['out'] if Number == type(arguments[1]) and call_utilities.is_iterable( arguments[0]): c_t = get_contained_elements_type(localization, arguments[0]) if not 'out' in dvar.keys(): return call_utilities.create_numpy_array(c_t) else: set_contained_elements_type(localization, dvar['out'], c_t) return dvar['out'] if call_utilities.is_iterable( arguments[0]) and call_utilities.is_iterable(arguments[1]): if call_utilities.get_dimensions( localization, arguments[0]) == 1 and call_utilities.get_dimensions( localization, arguments[1]) == 1: return call_utilities.cast_to_greater_numpy_type( call_utilities.get_inner_type(localization, arguments[0]), call_utilities.get_inner_type(localization, arguments[1])) typ = call_utilities.cast_to_greater_numpy_type( call_utilities.get_inner_type(localization, arguments[0]), call_utilities.get_inner_type(localization, arguments[1])) for i in range( call_utilities.get_dimensions(localization, arguments[0])): typ = call_utilities.create_numpy_array(typ) if not 'out' in dvar.keys(): return typ else: set_contained_elements_type( localization, dvar['out'], get_contained_elements_type(localization, typ)) return dvar['out'] return arguments[0]
def divide(localization, proxy_obj, arguments, func_name='divide'): dvar = call_utilities.parse_varargs_and_kwargs( localization, arguments, ['out', 'where'], { 'out': numpy.ndarray, 'where': numpy.ndarray, }, func_name, 2) if isinstance(dvar, StypyTypeError): return dvar if Number == type(arguments[0]) and Number == type(arguments[1]): return call_utilities.cast_to_greater_numpy_type( arguments[0], arguments[1]) try: dims = 0 if call_utilities.is_iterable(arguments[0]): param0 = call_utilities.get_inner_type(localization, arguments[0]) dims = call_utilities.get_dimensions(localization, arguments[0]) else: param0 = arguments[0] if call_utilities.is_iterable(arguments[1]): param1 = call_utilities.get_inner_type(localization, arguments[1]) temp = call_utilities.get_dimensions(localization, arguments[1]) if temp > dims: dims = temp else: param1 = arguments[1] if dims > 0: ret = call_utilities.create_numpy_array_n_dimensions( call_utilities.cast_to_greater_numpy_type(param0, param1), dims) else: ret = call_utilities.cast_to_greater_numpy_type(param0, param1) except Exception as ex: return StypyTypeError(localization, str(ex)) if 'out' in dvar.keys(): set_contained_elements_type(localization, dvar['out'], ret) return dvar['out'] return ret
def zeros_like(localization, proxy_obj, arguments): dvar = call_utilities.parse_varargs_and_kwargs(localization, arguments, ['dtype', 'order', 'subok'], { 'dtype': type, 'order': Str, 'subok': bool, }, 'ones') if isinstance(dvar, StypyTypeError): return dvar if 'dtype' in dvar.keys(): dtype = dvar['dtype'] else: dtype = None t = call_utilities.check_possible_values(dvar, 'order', ['C', 'F', 'A', 'K']) if isinstance(t, StypyTypeError): return t if Number == type(arguments[0]): return call_utilities.create_numpy_array(arguments[0], dtype=dtype) else: dims = call_utilities.get_dimensions(localization, arguments[0]) typ = call_utilities.create_numpy_array_n_dimensions(call_utilities.get_inner_type(localization, arguments[0]), dims, dtype=dtype) return typ
def trace(localization, proxy_obj, arguments): dvar = call_utilities.parse_varargs_and_kwargs( localization, arguments, ['offset', 'axis1', ' axis2', 'dtype', 'out'], { 'offset': Integer, 'axis1': Integer, 'axis2': Integer, 'dtype': type, 'out': numpy.ndarray, }, 'trace') if isinstance(dvar, StypyTypeError): return dvar dim = call_utilities.get_dimensions(localization, arguments[0]) if dim == 1: return call_utilities.cast_to_numpy_type( get_contained_elements_type(localization, arguments[0])) else: ret = call_utilities.create_numpy_array( call_utilities.get_inner_type(localization, arguments[0])) if 'out' in dvar.keys(): if dim == 1 or not (call_utilities.get_dimensions( localization, dvar['out']) == 1): return StypyTypeError( localization, "Wrong dimensions of out parameter in trace call") set_contained_elements_type( localization, dvar['out'], get_contained_elements_type(localization, arguments[0])) return dvar['out'] return ret
def reciprocal(localization, proxy_obj, arguments, func_name='reciprocal'): dvar = call_utilities.parse_varargs_and_kwargs( localization, arguments, ['out', 'where'], { 'out': numpy.ndarray, 'where': numpy.ndarray, }, func_name) if isinstance(dvar, StypyTypeError): return dvar if Number == type(arguments[0]): ret = call_utilities.cast_to_numpy_type(numpy.float64()) else: try: ret = call_utilities.create_numpy_array_n_dimensions( call_utilities.cast_to_numpy_type( call_utilities.get_inner_type(localization, arguments[0])), call_utilities.get_dimensions(localization, arguments[0])) except Exception as ex: return StypyTypeError(localization, str(ex)) if 'out' in dvar.keys(): set_contained_elements_type(localization, dvar['out'], ret) return dvar['out'] return ret
def __div__(localization, proxy_obj, arguments, func_name='__div__'): # dvar = call_utilities.parse_varargs_and_kwargs(localization, arguments, ['out', 'where'], { # 'out': numpy.ndarray, # 'where': numpy.ndarray, # }, func_name, 2) # # if isinstance(dvar, StypyTypeError): # return dvar try: dims = 0 if call_utilities.is_iterable(arguments[0]): param0 = call_utilities.get_inner_type(localization, arguments[0]) dims = call_utilities.get_dimensions(localization, arguments[0]) else: param0 = arguments[0] if dims > 0: ret = call_utilities.create_numpy_array_n_dimensions( call_utilities.cast_to_numpy_type(param0), dims) else: ret = call_utilities.create_numpy_array(call_utilities.cast_to_numpy_type(param0)) except Exception as ex: return StypyTypeError(localization, str(ex)) # if 'out' in dvar.keys(): # set_contained_elements_type(localization, dvar['out'], # ret) # return dvar['out'] return ret
def sum(localization, proxy_obj, arguments): if Number == type(arguments[0]): return call_utilities.cast_to_numpy_type(arguments[0]) dvar = call_utilities.parse_varargs_and_kwargs( localization, arguments, ['axis', 'dtype', 'out'], { 'axis': [ types.NoneType, Integer, IterableDataStructureWithTypedElements(Integer) ], 'dtype': type, 'out': numpy.ndarray, 'keepdims': bool }, 'sum') if isinstance(dvar, StypyTypeError): return dvar if 'out' in dvar.keys(): set_contained_elements_type( localization, dvar['out'], get_contained_elements_type(localization, arguments[0])) return dvar['out'] if 'axis' in dvar.keys(): return call_utilities.create_numpy_array( get_contained_elements_type(localization, arguments[0])) return call_utilities.cast_to_numpy_type( call_utilities.get_inner_type(localization, arguments[0]))
def det(localization, proxy_obj, arguments): param0 = call_utilities.get_inner_type(localization, arguments[0]) dims = call_utilities.get_dimensions(localization, arguments[0]) if dims == 1: return call_utilities.cast_to_numpy_type(param0) else: return call_utilities.create_numpy_array_n_dimensions( call_utilities.cast_to_numpy_type(param0), dims)
def einsum(localization, proxy_obj, arguments): if isinstance(arguments[-1], dict): if Str == type(arguments[0]): arg_num = 2 else: arg_num = 1 dvar = call_utilities.parse_varargs_and_kwargs(localization, arguments, ['out', 'dtype', 'order', 'casting', 'optimize'], { 'out': IterableDataStructure, 'dtype': type, 'order': Str, 'casting': Str, 'optimize': [bool, Str] }, 'einsum', arg_num) if isinstance(dvar, StypyTypeError): return dvar val_temp = call_utilities.check_possible_values(dvar, 'order', ['C', 'F', 'A', 'K']) if isinstance(val_temp, StypyTypeError): return val_temp val_temp = call_utilities.check_possible_values(dvar, 'casting', ['no', 'equiv', 'safe', 'same_kind', 'unsafe']) if isinstance(val_temp, StypyTypeError): return val_temp val_temp = call_utilities.check_possible_values(dvar, 'optimize', ['greedy', 'optimal', False, True]) if isinstance(val_temp, StypyTypeError): return val_temp arguments = arguments[:-1] else: dvar = dict() typ = None if Str == type(arguments[0]): arg_list = arguments[1:] if Number == type(arguments[1]) and 'out' in dvar: return dvar['out'] else: arg_list = arguments for arg in arg_list: if call_utilities.is_iterable(arg): typ_temp = call_utilities.cast_to_numpy_type(call_utilities.get_inner_type(localization, arg)) typ = call_utilities.cast_to_greater_numpy_type(typ, typ_temp) union = UnionType.add(typ, call_utilities.create_numpy_array(DynamicType)) if 'out' in dvar: set_contained_elements_type(localization, dvar['out'], DynamicType) return call_utilities.create_numpy_array(DynamicType) return union
def hstack(localization, proxy_obj, arguments): union = None for arg in arguments: if call_utilities.is_iterable(arg): union = UnionType.add( union, call_utilities.get_inner_type(localization, arg)) else: return StypyTypeError( localization, "A non-iterable parameter {0} was passed to the hstack function" .format(str(arg))) return call_utilities.create_numpy_array(union)
def rollaxis(localization, proxy_obj, arguments): dvar = call_utilities.parse_varargs_and_kwargs(localization, arguments, ['shift', 'axis'], { 'shift': [Integer, IterableDataStructureWithTypedElements(Integer)], 'axis': [Integer, IterableDataStructureWithTypedElements(Integer)], }, 'rollaxis', 2) if isinstance(dvar, StypyTypeError): return dvar t = call_utilities.get_inner_type(localization, arguments[0]) if not Number == type(t): return StypyTypeError(localization, "The contents of the passed array are not numeric") return arguments[0]
def repeat(localization, proxy_obj, arguments): dvar = call_utilities.parse_varargs_and_kwargs(localization, arguments, ['repeats', 'axis'], { 'repeats': [Integer, IterableDataStructureWithTypedElements(Integer)], 'axis': Integer, }, 'repeat') if isinstance(dvar, StypyTypeError): return dvar r = TypeWrapper.get_wrapper_of(proxy_obj.__self__) if 'axis' in dvar.keys(): return r else: typ = call_utilities.get_inner_type(localization, r) return call_utilities.create_numpy_array(typ)
def unpackbits(localization, proxy_obj, arguments): dvar = call_utilities.parse_varargs_and_kwargs(localization, arguments, ['axis'], { 'axis': Integer, }, 'unpackbits') if isinstance(dvar, StypyTypeError): return dvar if 'axis' in dvar.keys(): contained = call_utilities.get_contained_elements_type( localization, arguments[0]) else: contained = call_utilities.get_inner_type(localization, arguments[0]) return call_utilities.create_numpy_array(contained)
def atleast_2d(localization, proxy_obj, arguments): rets = list() for arg in arguments: if call_utilities.is_iterable(arg): if call_utilities.get_dimensions(localization, arg) >= 2: rets.append(arg) else: rets.append( call_utilities.create_numpy_array_n_dimensions( call_utilities.get_inner_type(localization, arg), 2)) else: return StypyTypeError( localization, "A non-iterable parameter {0} was passed to the atleast_2d function" .format(str(arg))) if len(rets) == 1: return rets[0] return tuple(rets)
def bitwise_and(localization, proxy_obj, arguments, func_name='bitwise_and'): dvar = call_utilities.parse_varargs_and_kwargs( localization, arguments, ['where'], { 'where': [Str, IterableDataStructureWithTypedElements(bool)], }, func_name, 2) if isinstance(dvar, StypyTypeError): return dvar if call_utilities.is_iterable( arguments[0]) and call_utilities.is_iterable(arguments[1]): if not call_utilities.check_possible_types( call_utilities.get_inner_type(localization, arguments[0]), [bool, Integer]) or not call_utilities.check_possible_types( call_utilities.get_inner_type(localization, arguments[1]), [bool, Integer]): return StypyTypeError( localization, " ufunc '" + func_name + "' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''" ) if call_utilities.is_numpy_array(arguments[0]): return call_utilities.create_numpy_array( call_utilities.get_inner_type(localization, arguments[0])) if call_utilities.is_numpy_array(arguments[1]): return call_utilities.create_numpy_array( call_utilities.get_inner_type(localization, arguments[1])) return arguments[0] else: if call_utilities.is_iterable( arguments[0]) and not call_utilities.is_iterable( arguments[1]): if not call_utilities.check_possible_types( call_utilities.get_inner_type(localization, arguments[0]), [bool, Integer ]) or not call_utilities.check_possible_types( arguments[1], [bool, Integer]): return StypyTypeError( localization, " ufunc '" + func_name + "' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''" ) if call_utilities.is_numpy_array(arguments[0]): return call_utilities.create_numpy_array( call_utilities.get_inner_type(localization, arguments[0])) return arguments[0] else: if not call_utilities.is_iterable( arguments[0]) and call_utilities.is_iterable( arguments[1]): if not call_utilities.check_possible_types( call_utilities.get_inner_type( localization, arguments[1]), [bool, Integer ]) or not call_utilities.check_possible_types( arguments[0], [bool, Integer]): return StypyTypeError( localization, " ufunc '" + func_name + "' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''" ) if call_utilities.is_numpy_array(arguments[1]): return call_utilities.create_numpy_array( call_utilities.get_inner_type( localization, arguments[1])) return arguments[1] else: return arguments[0]