Coverage for /var/srv/projects/api.amasfac.comuna18.com/tmp/venv/lib/python3.9/site-packages/pandas/core/internals/ops.py: 22%

62 statements  

« prev     ^ index     » next       coverage.py v6.4.4, created at 2023-07-17 14:22 -0600

1from __future__ import annotations 

2 

3from typing import ( 

4 TYPE_CHECKING, 

5 Iterator, 

6 NamedTuple, 

7) 

8 

9from pandas._typing import ArrayLike 

10 

11if TYPE_CHECKING: 11 ↛ 12line 11 didn't jump to line 12, because the condition on line 11 was never true

12 from pandas._libs.internals import BlockPlacement 

13 

14 from pandas.core.internals.blocks import Block 

15 from pandas.core.internals.managers import BlockManager 

16 

17 

18class BlockPairInfo(NamedTuple): 

19 lvals: ArrayLike 

20 rvals: ArrayLike 

21 locs: BlockPlacement 

22 left_ea: bool 

23 right_ea: bool 

24 rblk: Block 

25 

26 

27def _iter_block_pairs( 

28 left: BlockManager, right: BlockManager 

29) -> Iterator[BlockPairInfo]: 

30 # At this point we have already checked the parent DataFrames for 

31 # assert rframe._indexed_same(lframe) 

32 

33 for blk in left.blocks: 

34 locs = blk.mgr_locs 

35 blk_vals = blk.values 

36 

37 left_ea = blk_vals.ndim == 1 

38 

39 rblks, _ = right._slice_take_blocks_ax0(locs.indexer, only_slice=True) 

40 

41 # Assertions are disabled for performance, but should hold: 

42 # if left_ea: 

43 # assert len(locs) == 1, locs 

44 # assert len(rblks) == 1, rblks 

45 # assert rblks[0].shape[0] == 1, rblks[0].shape 

46 

47 for rblk in rblks: 

48 right_ea = rblk.values.ndim == 1 

49 

50 lvals, rvals = _get_same_shape_values(blk, rblk, left_ea, right_ea) 

51 info = BlockPairInfo(lvals, rvals, locs, left_ea, right_ea, rblk) 

52 yield info 

53 

54 

55def operate_blockwise( 

56 left: BlockManager, right: BlockManager, array_op 

57) -> BlockManager: 

58 # At this point we have already checked the parent DataFrames for 

59 # assert rframe._indexed_same(lframe) 

60 

61 res_blks: list[Block] = [] 

62 for lvals, rvals, locs, left_ea, right_ea, rblk in _iter_block_pairs(left, right): 

63 res_values = array_op(lvals, rvals) 

64 if left_ea and not right_ea and hasattr(res_values, "reshape"): 

65 res_values = res_values.reshape(1, -1) 

66 nbs = rblk._split_op_result(res_values) 

67 

68 # Assertions are disabled for performance, but should hold: 

69 # if right_ea or left_ea: 

70 # assert len(nbs) == 1 

71 # else: 

72 # assert res_values.shape == lvals.shape, (res_values.shape, lvals.shape) 

73 

74 _reset_block_mgr_locs(nbs, locs) 

75 

76 res_blks.extend(nbs) 

77 

78 # Assertions are disabled for performance, but should hold: 

79 # slocs = {y for nb in res_blks for y in nb.mgr_locs.as_array} 

80 # nlocs = sum(len(nb.mgr_locs.as_array) for nb in res_blks) 

81 # assert nlocs == len(left.items), (nlocs, len(left.items)) 

82 # assert len(slocs) == nlocs, (len(slocs), nlocs) 

83 # assert slocs == set(range(nlocs)), slocs 

84 

85 new_mgr = type(right)(tuple(res_blks), axes=right.axes, verify_integrity=False) 

86 return new_mgr 

87 

88 

89def _reset_block_mgr_locs(nbs: list[Block], locs): 

90 """ 

91 Reset mgr_locs to correspond to our original DataFrame. 

92 """ 

93 for nb in nbs: 

94 nblocs = locs[nb.mgr_locs.indexer] 

95 nb.mgr_locs = nblocs 

96 # Assertions are disabled for performance, but should hold: 

97 # assert len(nblocs) == nb.shape[0], (len(nblocs), nb.shape) 

98 # assert all(x in locs.as_array for x in nb.mgr_locs.as_array) 

99 

100 

101def _get_same_shape_values( 

102 lblk: Block, rblk: Block, left_ea: bool, right_ea: bool 

103) -> tuple[ArrayLike, ArrayLike]: 

104 """ 

105 Slice lblk.values to align with rblk. Squeeze if we have EAs. 

106 """ 

107 lvals = lblk.values 

108 rvals = rblk.values 

109 

110 # Require that the indexing into lvals be slice-like 

111 assert rblk.mgr_locs.is_slice_like, rblk.mgr_locs 

112 

113 # TODO(EA2D): with 2D EAs only this first clause would be needed 

114 if not (left_ea or right_ea): 

115 # error: No overload variant of "__getitem__" of "ExtensionArray" matches 

116 # argument type "Tuple[Union[ndarray, slice], slice]" 

117 lvals = lvals[rblk.mgr_locs.indexer, :] # type: ignore[call-overload] 

118 assert lvals.shape == rvals.shape, (lvals.shape, rvals.shape) 

119 elif left_ea and right_ea: 

120 assert lvals.shape == rvals.shape, (lvals.shape, rvals.shape) 

121 elif right_ea: 

122 # lvals are 2D, rvals are 1D 

123 

124 # error: No overload variant of "__getitem__" of "ExtensionArray" matches 

125 # argument type "Tuple[Union[ndarray, slice], slice]" 

126 lvals = lvals[rblk.mgr_locs.indexer, :] # type: ignore[call-overload] 

127 assert lvals.shape[0] == 1, lvals.shape 

128 lvals = lvals[0, :] 

129 else: 

130 # lvals are 1D, rvals are 2D 

131 assert rvals.shape[0] == 1, rvals.shape 

132 # error: No overload variant of "__getitem__" of "ExtensionArray" matches 

133 # argument type "Tuple[int, slice]" 

134 rvals = rvals[0, :] # type: ignore[call-overload] 

135 

136 return lvals, rvals 

137 

138 

139def blockwise_all(left: BlockManager, right: BlockManager, op) -> bool: 

140 """ 

141 Blockwise `all` reduction. 

142 """ 

143 for info in _iter_block_pairs(left, right): 

144 res = op(info.lvals, info.rvals) 

145 if not res: 

146 return False 

147 return True