[英]Converting rows to wide columns based on duplicated ids in another column in pandas
My question is similar to this , this , and this questions.我的问题类似于this 、 this和this问题。
But still cannot resolve it.但仍然无法解决。
I have a dataframe with duplicated ids我有一个带有重复 ID 的 dataframe
ID Publication_type
1 Journal
1 Clinical study
1 Guideline
2 Journal
2 Letter
I want to make it wide, but I do not know how many publication type will I have - maybe 2, maybe 20. Thus, I do not know how many columns wide will I need.我想让它变宽,但我不知道我会有多少种出版物类型——也许是 2,也许是 20。因此,我不知道我需要多少列宽。 The max size of wide columns for publication_type
must be not be more than the number of types for each id. publication_type
的宽列的最大大小不得超过每个 id 的类型数。
Expected output预期 output
ID Publication_type1 Publication_type2 Publication_type 3 etc
1 Journal Clinical Study Guideline
2 Journal Letter NaN
For now I do not need to put the same publication type into the same column.现在我不需要将相同的发布类型放入同一列。 I do not need all articles in the same column.我不需要同一列中的所有文章。 Thanks!谢谢!
You can group by ID
, aggregate via list
, and then create a new DataFrame from the results:您可以按ID
分组,通过list
聚合,然后从结果中创建一个新的 DataFrame:
col = 'Publication_type'
new_df = pd.DataFrame(df.groupby('ID')[col].agg(lambda x: x.tolist()).tolist()).replace({None: np.nan})
new_df.columns = [f'{col}{i}' for i in new_df.columns + 1]
new_df['ID'] = df['ID'].drop_duplicates().reset_index(drop=True)
Output: Output:
>>> df
Publication_type1 Publication_type2 Publication_type3 ID
0 Journal Clinical-study Guideline 1
1 Journal Letter NaN 2
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.