df数据表：

汇总每个name对应的frequency和CTR，去掉列名CTR，列名用具体的name替代，结果如下（示例为结果的一部分）：

参考答案：

np.random.seed(0)
df = pd.DataFrame({"frequency":np.random.randint(1,20,8)
              ,"CTR":np.random.uniform(0,1,8)
              ,"name":["0234s3","023dfr","34s231","a34d21","ss034q","34s231","0234s3","023dfr"]})
gp = df.groupby("name").apply(lambda x : x.sort_values("CTR"))
gp.columns = ['frequency', 'ctr', 'name1']
name = gp.pop("name1")
name.value_counts().index
df1 = pd.DataFrame()
for i in name:
    df0 = gp.loc[i].reset_index().drop(columns = ["index"])
    df0.columns = ["frequency",i]
    df1 = pd.concat([df1,df0],axis=1)
df1

结果示例：