Python base.W_NDimArray类代码示例

 def execute(self, interp):
     w_lhs = self.lhs.execute(interp)
     if isinstance(self.rhs, SliceConstant):
         w_rhs = self.rhs.wrap(interp.space)
     else:
         w_rhs = self.rhs.execute(interp)
     if not isinstance(w_lhs, W_NDimArray):
         # scalar
         dtype = get_dtype_cache(interp.space).w_float64dtype
         w_lhs = W_NDimArray.new_scalar(interp.space, dtype, w_lhs)
     assert isinstance(w_lhs, W_NDimArray)
     if self.name == '+':
         w_res = w_lhs.descr_add(interp.space, w_rhs)
     elif self.name == '*':
         w_res = w_lhs.descr_mul(interp.space, w_rhs)
     elif self.name == '-':
         w_res = w_lhs.descr_sub(interp.space, w_rhs)
     elif self.name == '->':
         if isinstance(w_rhs, FloatObject):
             w_rhs = IntObject(int(w_rhs.floatval))
         assert isinstance(w_lhs, W_NDimArray)
         w_res = w_lhs.descr_getitem(interp.space, w_rhs)
     else:
         raise NotImplementedError
     if (not isinstance(w_res, W_NDimArray) and
         not isinstance(w_res, interp_boxes.W_GenericBox)):
         dtype = get_dtype_cache(interp.space).w_float64dtype
         w_res = W_NDimArray.new_scalar(interp.space, dtype, w_res)
     return w_res

def frombuffer(space, w_buffer, w_dtype=None, count=-1, offset=0):
    dtype = space.interp_w(descriptor.W_Dtype,
        space.call_function(space.gettypefor(descriptor.W_Dtype), w_dtype))
    if dtype.elsize == 0:
        raise oefmt(space.w_ValueError, "itemsize cannot be zero in type")

    try:
        buf = _getbuffer(space, w_buffer)
    except OperationError as e:
        if not e.match(space, space.w_TypeError):
            raise
        w_buffer = space.call_method(w_buffer, '__buffer__', 
                                    space.newint(space.BUF_FULL_RO))
        buf = _getbuffer(space, w_buffer)

    ts = buf.getlength()
    if offset < 0 or offset > ts:
        raise oefmt(space.w_ValueError,
                    "offset must be non-negative and no greater than "
                    "buffer length (%d)", ts)

    s = ts - offset
    if offset:
        buf = SubBuffer(buf, offset, s)

    n = count
    itemsize = dtype.elsize
    assert itemsize > 0
    if n < 0:
        if s % itemsize != 0:
            raise oefmt(space.w_ValueError,
                        "buffer size must be a multiple of element size")
        n = s / itemsize
    else:
        if s < n * itemsize:
            raise oefmt(space.w_ValueError,
                        "buffer is smaller than requested size")

    try:
        storage = buf.get_raw_address()
    except ValueError:
        a = W_NDimArray.from_shape(space, [n], dtype=dtype)
        loop.fromstring_loop(space, a, dtype, itemsize, buf.as_str())
        return a
    else:
        writable = not buf.readonly
    return W_NDimArray.from_shape_and_storage(space, [n], storage, storage_bytes=s,
                                dtype=dtype, w_base=w_buffer, writable=writable)

 def apply(self, space, orig_arr):
     arr = orig_arr.implementation
     ofs, subdtype = arr.dtype.fields[self.name]
     # strides backstrides are identical, ofs only changes start
     return W_NDimArray.new_slice(space, arr.start + ofs, arr.get_strides(),
                                  arr.get_backstrides(),
                                  arr.shape, arr, orig_arr, subdtype)

def zeros(space, w_shape, w_dtype=None, w_order=None):
    dtype = space.interp_w(descriptor.W_Dtype,
        space.call_function(space.gettypefor(descriptor.W_Dtype), w_dtype))
    if dtype.is_str_or_unicode() and dtype.elsize < 1:
        dtype = descriptor.variable_dtype(space, dtype.char + '1')
    shape = shape_converter(space, w_shape, dtype)
    return W_NDimArray.from_shape(space, shape, dtype=dtype)

def accumulate(space, func, w_arr, axis, calc_dtype, w_out, identity):
    out_iter, out_state = w_out.create_iter()
    arr_shape = w_arr.get_shape()
    temp_shape = arr_shape[:axis] + arr_shape[axis + 1:]
    temp = W_NDimArray.from_shape(space, temp_shape, calc_dtype, w_instance=w_arr)
    temp_iter = AxisIter(temp.implementation, w_arr.get_shape(), axis)
    temp_state = temp_iter.reset()
    arr_iter, arr_state = w_arr.create_iter()
    arr_iter.track_index = False
    if identity is not None:
        identity = identity.convert_to(space, calc_dtype)
    shapelen = len(arr_shape)
    while not out_iter.done(out_state):
        accumulate_driver.jit_merge_point(shapelen=shapelen, func=func,
                                          calc_dtype=calc_dtype)
        w_item = arr_iter.getitem(arr_state).convert_to(space, calc_dtype)
        arr_state = arr_iter.next(arr_state)

        out_indices = out_iter.indices(out_state)
        if out_indices[axis] == 0:
            if identity is not None:
                w_item = func(calc_dtype, identity, w_item)
        else:
            cur_value = temp_iter.getitem(temp_state)
            w_item = func(calc_dtype, cur_value, w_item)

        out_iter.setitem(out_state, w_item)
        out_state = out_iter.next(out_state)
        temp_iter.setitem(temp_state, w_item)
        temp_state = temp_iter.next(temp_state)
    return w_out

def numpify(space, w_object):
    """Convert the object to a W_NumpyObject"""
    # XXX: code duplication with _array()
    from pypy.module.micronumpy import strides
    if isinstance(w_object, W_NumpyObject):
        return w_object
    # for anything that isn't already an array, try __array__ method first
    w_array = try_array_method(space, w_object)
    if w_array is not None:
        return w_array

    shape, elems_w = strides.find_shape_and_elems(space, w_object, None)
    dtype = find_dtype_for_seq(space, elems_w, None)
    if dtype is None:
        dtype = descriptor.get_dtype_cache(space).w_float64dtype
    elif dtype.is_str_or_unicode() and dtype.elsize < 1:
        # promote S0 -> S1, U0 -> U1
        dtype = descriptor.variable_dtype(space, dtype.char + '1')

    if len(elems_w) == 1:
        return dtype.coerce(space, elems_w[0])
    else:
        w_arr = W_NDimArray.from_shape(space, shape, dtype)
        loop.assign(space, w_arr, elems_w)
        return w_arr

def new_view(space, w_arr, chunks):
    arr = w_arr.implementation
    r = calculate_slice_strides(space, arr.shape, arr.start, arr.get_strides(),
                                arr.get_backstrides(), chunks)
    shape, start, strides, backstrides = r
    return W_NDimArray.new_slice(space, start, strides[:], backstrides[:],
                                 shape[:], arr, w_arr)

def call2(space, shape, func, calc_dtype, res_dtype, w_lhs, w_rhs, out):
    # handle array_priority
    # w_lhs and w_rhs could be of different ndarray subtypes. Numpy does:
    # 1. if __array_priorities__ are equal and one is an ndarray and the
    #        other is a subtype,  flip the order
    # 2. elif rhs.__array_priority__ is higher, flip the order
    # Now return the subtype of the first one

    w_ndarray = space.gettypefor(W_NDimArray)
    lhs_type = space.type(w_lhs)
    rhs_type = space.type(w_rhs)
    lhs_for_subtype = w_lhs
    rhs_for_subtype = w_rhs
    #it may be something like a FlatIter, which is not an ndarray
    if not space.is_true(space.issubtype(lhs_type, w_ndarray)):
        lhs_type = space.type(w_lhs.base)
        lhs_for_subtype = w_lhs.base
    if not space.is_true(space.issubtype(rhs_type, w_ndarray)):
        rhs_type = space.type(w_rhs.base)
        rhs_for_subtype = w_rhs.base
    if space.is_w(lhs_type, w_ndarray) and not space.is_w(rhs_type, w_ndarray):
        lhs_for_subtype = rhs_for_subtype

    # TODO handle __array_priorities__ and maybe flip the order

    if w_lhs.get_size() == 1:
        w_left = w_lhs.get_scalar_value().convert_to(space, calc_dtype)
        left_iter = left_state = None
    else:
        w_left = None
        left_iter, left_state = w_lhs.create_iter(shape)
        left_iter.track_index = False

    if w_rhs.get_size() == 1:
        w_right = w_rhs.get_scalar_value().convert_to(space, calc_dtype)
        right_iter = right_state = None
    else:
        w_right = None
        right_iter, right_state = w_rhs.create_iter(shape)
        right_iter.track_index = False

    if out is None:
        out = W_NDimArray.from_shape(space, shape, res_dtype,
                                     w_instance=lhs_for_subtype)
    out_iter, out_state = out.create_iter(shape)
    shapelen = len(shape)
    while not out_iter.done(out_state):
        call2_driver.jit_merge_point(shapelen=shapelen, func=func,
                                     calc_dtype=calc_dtype, res_dtype=res_dtype)
        if left_iter:
            w_left = left_iter.getitem(left_state).convert_to(space, calc_dtype)
            left_state = left_iter.next(left_state)
        if right_iter:
            w_right = right_iter.getitem(right_state).convert_to(space, calc_dtype)
            right_state = right_iter.next(right_state)
        out_iter.setitem(out_state, func(calc_dtype, w_left, w_right).convert_to(
            space, res_dtype))
        out_state = out_iter.next(out_state)
    return out

 def apply(self, space, orig_arr):
     arr = orig_arr.implementation
     shape = self.extend_shape(arr.shape)
     r = calculate_slice_strides(arr.shape, arr.start, arr.get_strides(),
                                 arr.get_backstrides(), self.l)
     _, start, strides, backstrides = r
     return W_NDimArray.new_slice(space, start, strides[:], backstrides[:],
                                  shape[:], arr, orig_arr)

 def swapaxes(self, space, orig_arr, axis1, axis2):
     shape = self.get_shape()[:]
     strides = self.get_strides()[:]
     backstrides = self.get_backstrides()[:]
     shape[axis1], shape[axis2] = shape[axis2], shape[axis1]
     strides[axis1], strides[axis2] = strides[axis2], strides[axis1]
     backstrides[axis1], backstrides[axis2] = backstrides[axis2], backstrides[axis1]
     return W_NDimArray.new_slice(space, self.start, strides,
                                  backstrides, shape, self, orig_arr)

 def set_shape(self, space, orig_array, new_shape):
     if not new_shape:
         return self
     if support.product(new_shape) == 1:
         arr = W_NDimArray.from_shape(space, new_shape, self.dtype)
         arr_iter = arr.create_iter(new_shape)
         arr_iter.setitem(self.value)
         return arr.implementation
     raise OperationError(space.w_ValueError, space.wrap("total size of the array must be unchanged"))

 def nonzero(self, space, index_type):
     s = loop.count_all_true_concrete(self)
     box = index_type.itemtype.box
     nd = len(self.get_shape()) or 1
     w_res = W_NDimArray.from_shape(space, [s, nd], index_type)
     loop.nonzero(w_res, self, box)
     w_res = w_res.implementation.swapaxes(space, w_res, 0, 1)
     l_w = [w_res.descr_getitem(space, space.wrap(d)) for d in range(nd)]
     return space.newtuple(l_w)

def concatenate(space, w_args, w_axis=None):
    args_w = space.listview(w_args)
    if len(args_w) == 0:
        raise oefmt(space.w_ValueError, "need at least one array to concatenate")
    args_w = [convert_to_array(space, w_arg) for w_arg in args_w]
    if w_axis is None:
        w_axis = space.wrap(0)
    if space.is_none(w_axis):
        args_w = [w_arg.reshape(space,
                                space.newlist([w_arg.descr_get_size(space)]),
                                w_arg.get_order())
                  for w_arg in args_w]
        w_axis = space.wrap(0)
    dtype = args_w[0].get_dtype()
    shape = args_w[0].get_shape()[:]
    ndim = len(shape)
    if ndim == 0:
        raise oefmt(space.w_ValueError,
                    "zero-dimensional arrays cannot be concatenated")
    axis = space.int_w(w_axis)
    orig_axis = axis
    if axis < 0:
        axis = ndim + axis
    if ndim == 1 and axis != 0:
        axis = 0
    if axis < 0 or axis >= ndim:
        raise oefmt(space.w_IndexError, "axis %d out of bounds [0, %d)",
                    orig_axis, ndim)
    for arr in args_w[1:]:
        if len(arr.get_shape()) != ndim:
            raise oefmt(space.w_ValueError,
                        "all the input arrays must have same number of "
                        "dimensions")
        for i, axis_size in enumerate(arr.get_shape()):
            if i == axis:
                shape[i] += axis_size
            elif axis_size != shape[i]:
                raise oefmt(space.w_ValueError,
                            "all the input array dimensions except for the "
                            "concatenation axis must match exactly")

    dtype = find_result_type(space, args_w, [])
    # concatenate does not handle ndarray subtypes, it always returns a ndarray
    res = W_NDimArray.from_shape(space, shape, dtype, NPY.CORDER)
    chunks = [Chunk(0, i, 1, i) for i in shape]
    axis_start = 0
    for arr in args_w:
        if arr.get_shape()[axis] == 0:
            continue
        chunks[axis] = Chunk(axis_start, axis_start + arr.get_shape()[axis], 1,
                             arr.get_shape()[axis])
        view = new_view(space, res, chunks)
        view.implementation.setslice(space, arr)
        axis_start += arr.get_shape()[axis]
    return res

def empty_like(space, w_a, w_dtype=None, w_order=None, subok=True):
    w_a = convert_to_array(space, w_a)
    if space.is_none(w_dtype):
        dtype = w_a.get_dtype()
    else:
        dtype = space.interp_w(descriptor.W_Dtype,
            space.call_function(space.gettypefor(descriptor.W_Dtype), w_dtype))
        if dtype.is_str_or_unicode() and dtype.elsize < 1:
            dtype = descriptor.variable_dtype(space, dtype.char + '1')
    return W_NDimArray.from_shape(space, w_a.get_shape(), dtype=dtype,
                                  w_instance=w_a if subok else None)

def tostring(space, arr):
    builder = StringBuilder()
    iter, state = arr.create_iter()
    w_res_str = W_NDimArray.from_shape(space, [1], arr.get_dtype(), order="C")
    itemsize = arr.get_dtype().elsize
    with w_res_str.implementation as storage:
        res_str_casted = rffi.cast(rffi.CArrayPtr(lltype.Char), support.get_storage_as_int(storage))
        while not iter.done(state):
            w_res_str.implementation.setitem(0, iter.getitem(state))
            for i in range(itemsize):
                builder.append(res_str_casted[i])
            state = iter.next(state)
        return builder.build()