pandas 数据帧ValueError:传递值的形状为(X,),索引暗示(X,Y) [英] Pandas Dataframe ValueError: Shape of passed values is (X, ), indices imply (X, Y)
问题描述
我收到错误,我不知道如何解决它。
I am getting an error and I'm not sure how to fix it.
以下内容似乎有效:
def random(row):
return [1,2,3,4]
df = pandas.DataFrame(np.random.randn(5, 4), columns=list('ABCD'))
df.apply(func = random, axis = 1)
和我的输出是:
[1,2,3,4]
[1,2,3,4]
[1,2,3,4]
[1,2,3,4]
但是,当我将其中一列更改为1或None的值时:
However, when I change one of the of the columns to a value such as 1 or None:
def random(row):
return [1,2,3,4]
df = pandas.DataFrame(np.random.randn(5, 4), columns=list('ABCD'))
df['E'] = 1
df.apply(func = random, axis = 1)
我收到错误:
ValueError: Shape of passed values is (5,), indices imply (5, 5)
我一直在摔跤现在几天,似乎没有任何工作。有趣的是,当我更改
I've been wrestling with this for a few days now and nothing seems to work. What is interesting is that when I change
def random(row):
return [1,2,3,4]
to
def random(row):
print [1,2,3,4]
$ b一切似乎都正常工作。
everything seems to work normally.
这个问题是提出这个问题,我觉得可能会令人困惑。
This question is a clearer way of asking this question, which I feel may have been confusing.
我的目标是计算每行的列表,然后创建一个列。
My goal is to compute a list for each row and then create a column out of that.
编辑:我最初是从一个数据框开始,一个列。我添加4列4个差异申请步骤,然后当我尝试添加另一列我得到这个错误。
I originally start with a dataframe that hase one column. I add 4 columns in 4 difference apply steps, and then when I try to add another column I get this error.
推荐答案
如果您的目标是向DataFrame添加新列,只需将函数写入返回标量值(不是列表)的函数,像这样:
If your goal is add new column to DataFrame, just write your function as function returning scalar value (not list), something like this:
>>> def random(row):
... return row.mean()
然后使用申请:
>>> df['new'] = df.apply(func = random, axis = 1)
>>> df
A B C D new
0 0.201143 -2.345828 -2.186106 -0.784721 -1.278878
1 -0.198460 0.544879 0.554407 -0.161357 0.184867
2 0.269807 1.132344 0.120303 -0.116843 0.351403
3 -1.131396 1.278477 1.567599 0.483912 0.549648
4 0.288147 0.382764 -0.840972 0.838950 0.167222
我不知道如果您的新列可能包含列表,但它无限可能包含元组((...)
而不是 [...]
):
I don't know if it possible for your new column to contain lists, but it deinitely possible to contain tuples ((...)
instead of [...]
):
>>> def random(row):
... return (1,2,3,4,5)
...
>>> df['new'] = df.apply(func = random, axis = 1)
>>> df
A B C D new
0 0.201143 -2.345828 -2.186106 -0.784721 (1, 2, 3, 4, 5)
1 -0.198460 0.544879 0.554407 -0.161357 (1, 2, 3, 4, 5)
2 0.269807 1.132344 0.120303 -0.116843 (1, 2, 3, 4, 5)
3 -1.131396 1.278477 1.567599 0.483912 (1, 2, 3, 4, 5)
4 0.288147 0.382764 -0.840972 0.838950 (1, 2, 3, 4, 5)
这篇关于 pandas 数据帧ValueError:传递值的形状为(X,),索引暗示(X,Y)的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!