明白了，就这么简单：

import pandas as pd

lis = {'a', 'b', 'c'}

output_df = pd.DataFrame()

for i in lis:
    get_data = download_data_from_csv(xxxxxxxxxxxx_i)
    # Assume get_data is a column array
    output_df[i] = get_data

以Series的形式收集数据，然后使用pd.concat合并它们，最后使用unstack合并第一层（名称为lis）以获得预期的输出：

# Dummy function to simulate downloaded data
def download_data_from_csv(i):
    """Simulate data"""
    d = {'a': [1, 6], 'b': [7, 10], 'c': [4, 9]}
    return d[i]

lis = ['a', 'b', 'c']  # Use a list not a set!
get_data = {i: pd.Series(download_data_from_csv(i)) for i in lis}
output_df = pd.concat(get_data).unstack(level=0)

输出量：

>>> output_df
   a   b  c
0  1   7  4
1  6  10  9