Python 将数据帧转换为嵌套JSON数组

我的df如下所示：

                dates  values
0 2020-01-01 00:15:00    25.7
1 2020-01-01 00:30:00    25.0
2 2020-01-01 00:45:00    24.6
3 2020-01-01 01:00:00    24.6
4 2020-01-01 01:15:00    25.0
5 2020-01-01 01:30:00    25.6
6 2020-01-01 01:45:00    26.2
7 2020-01-01 02:00:00    26.5
8 2020-01-01 02:15:00    26.3
9 2020-01-01 02:30:00    25.7

{
    "items": "[{\"dates\":1577837700,\"values\":25.7},{\"dates\":1577838600,\"values\":25.0},{\"dates\":1577839500,\"values\":24.6},{\"dates\":1577840400,\"values\":24.6},{\"dates\":1577841300,\"values\":25.0},{\"dates\":1577842200,\"values\":25.6},{\"dates\":1577843100,\"values\":26.2},{\"dates\":1577844000,\"values\":26.5},{\"dates\":1577844900,\"values\":26.3},{\"dates\":1577845800,\"values\":25.7}]" }

我确实：

df = df.to_json(orient='records', date_unit='s')
print({'items' : df})

它给我的输出如下：

                dates  values
0 2020-01-01 00:15:00    25.7
1 2020-01-01 00:30:00    25.0
2 2020-01-01 00:45:00    24.6
3 2020-01-01 01:00:00    24.6
4 2020-01-01 01:15:00    25.0
5 2020-01-01 01:30:00    25.6
6 2020-01-01 01:45:00    26.2
7 2020-01-01 02:00:00    26.5
8 2020-01-01 02:15:00    26.3
9 2020-01-01 02:30:00    25.7

{
    "items": "[{\"dates\":1577837700,\"values\":25.7},{\"dates\":1577838600,\"values\":25.0},{\"dates\":1577839500,\"values\":24.6},{\"dates\":1577840400,\"values\":24.6},{\"dates\":1577841300,\"values\":25.0},{\"dates\":1577842200,\"values\":25.6},{\"dates\":1577843100,\"values\":26.2},{\"dates\":1577844000,\"values\":26.5},{\"dates\":1577844900,\"values\":26.3},{\"dates\":1577845800,\"values\":25.7}]" }

我希望输出像这样

{
    "items": [[1577837700, 25.7],[1577838600,25.0],[1577839500,24.6],[1577840400,24.6],[1577841300,25.0],[1577842200,25.6],[1577843100,26.2],[1577844000,26.5],[1577844900,26.3],[1577845800,25.7]] }

也就是说，从代码的输出中，我希望所有记录都以

列表的形式出现，而不是
dict

有办法吗？
试试
orient=“values”
：


这只是将
.values
（或
.to_numpy
）输出转储为json。更多信息，请参阅