熊猫，无法连接数据帧

Question

我目前正在使用熊猫。 所以我尝试连接DataFrame，但它不起作用，所以我有一个问题。 代码如下

df # shape is (27796, 876)
genes_pca # shape is (27796, 50)
cells_pca # shape is (27796, 15)
# concat dataframe axis=1 result shape is (27796, 926)
df = pd.concat([df, genes_pca, cells_pca], axis=1)

所以，我收到了这个错误。

File "/opt/conda/lib/python3.7/site-packages/pandas/core/reshape/concat.py", line 287, in concat
moa-gpu_1  |     return op.get_result()
moa-gpu_1  |   File "/opt/conda/lib/python3.7/site-packages/pandas/core/reshape/concat.py", line 503, in get_result
moa-gpu_1  |     mgrs_indexers, self.new_axes, concat_axis=self.bm_axis, copy=self.copy,
moa-gpu_1  |   File "/opt/conda/lib/python3.7/site-packages/pandas/core/internals/concat.py", line 84, in concatenate_block_managers
moa-gpu_1  |     return BlockManager(blocks, axes)
moa-gpu_1  |   File "/opt/conda/lib/python3.7/site-packages/pandas/core/internals/managers.py", line 149, in __init__
moa-gpu_1  |     self._verify_integrity()
moa-gpu_1  |   File "/opt/conda/lib/python3.7/site-packages/pandas/core/internals/managers.py", line 326, in _verify_integrity
moa-gpu_1  |     raise construction_error(tot_items, block.shape[1:], self.axes)
moa-gpu_1  | ValueError: Shape of passed values is (39742, 941), indices imply (31778, 941)

我不知道您在此错误中遇到的 (39742, 941) 和 (31778, 941) 之类的数字是什么意思。

Answer 1

尝试在pd.concate 中使用ignore index 参数。此外，列将是 941 而不是 926。如果没有任何重叠的列

Answer 2

您是否尝试过重置索引？

df.reset_index(drop=True, inplace=True)
genes_pca.reset_index(drop=True, inplace=True)
cells_pca.reset_index(drop=True, inplace=True)
df = pd.concat([df, genes_pca, cells_pca], axis=1)

此外，检查数据帧中的重复索引值也是有意义的，例如

df.index.is_unique

如果存在重复，可以将其删除：

df.drop_duplicates(inplace=True)