pandas：转换数据帧集合时，缓冲区的维数错误（预期为1，得0）

Question

I have a dataframe with a column containing one-element sets. The one-elements are always of type str . I want to change this column's values to that one str inside each set.

My understanding of pandas so far has lead me to use the transform function on my dataframe. Because I only want to change one column, I call it on that column specifically.

I've tried map , transform , and apply (I can't(?) use applymap because the column is simply a series). Examples of what I've tried is this:

df['foo'].map(''.join)
df['foo'].transform(''.join)
df['foo'].apply(''.join)

this returns my column as a series, but doesn't change df 's foo column's values in-place. I then tried these:

df['foo'] = df['foo'].map(''.join)
df['foo'] = df['foo'].transform(''.join)
df['foo'] = df['foo'].apply(''.join)

swap out ''.join for lambda x: next(iter(x)) and the same error occurs.

some sample data from head(df) (thanks for the suggestion Yuca):

treatment             AAA        AAA        AAA  ...       BBB        BBB   foo                               
15025                 0.0        0.0        0.0  ...  126250.0   191730.0  {K00134}
6746            1523000.0  2374300.0  1394500.0  ...  536750.0  1179000.0  {K00134}
11097           1146200.0   423280.0   316280.0  ...       0.0   115880.0  {K00895}
3098             193000.0    57943.0    94483.0  ...       0.0   191530.0  {K01689}
14797             53168.0        0.0    65262.0  ...       0.0        0.0  {K00169}

[5 rows x 9 columns]

I think my understanding of apply , map , transform , etc. needs some work, but I just want to figure out how to do this and maybe why my method doesn't work. Thanks!

Answer 1

fix : as @Yuca mentioned, I had duplicate column names. This was because I had earlier removed 2/3 levels from the original MultiIndex assigned to df 's columns. When I re-assigned the columns with unique names, the lines with = worked! Thanks everyone for your help.