从 .json 文件生成的数据框，“日期”列被更改（显示图片）我该如何防止这种情况？

Question

我对 python 比较陌生，我被分配了一项数据科学任务，我必须为此选择一个 API（选择 global-warming.org），然后清理、解析数据并将其存储为 json 文件，然后将其加载到数据框中以进行进一步分析。然后我必须使用 matplotlib 等分析数据。我决定提取温室气体（CH4、CO2、N2O）的每个 API，将每个数据集缩减到 2020 年 3 月——现在，然后看看我是否可以分析大流行如何影响温室气体的产生。然而，我正在努力使用我存储的 json 文件将初始甲烷数据（每年每月采集一次）加载到一个干净的数据框中。如下图所示，“日期”列似乎与例如2020.4 格式转成这个格式：1970-01-01 00:33:40.700 ?

如果有人能告诉我如何解决这个问题，我会非常非常感激。如果有人对如何完成任务有任何一般性建议，我会欠你的，但我目前正被困在这个问题上，弄清楚它本身就是一个巨大的帮助，我相信这很简单我不见了……

非常感谢！

yearly_methane.json 文件示例行：

[{"date": "1983.7", "average": "1625.9", "trend": "1634.6", "averageUnc": "2.4", "trendUnc": "1.5"}, {"date": "1983.8", "average": "1628.1", "trend": "1635.3", "averageUnc": "2.9", "trendUnc": "1.4"}, {"date": "1983.9", "average": "1638.4", "trend": "1636.0", "averageUnc": "2.2", "trendUnc": "1.3"}, {"date": "1983.10", "average": "1644.8", "trend": "1636.7", "averageUnc": "1.4", "trendUnc": "1.3"}, {"date": "1983.11", "average": "1642.6", "trend": "1637.5", "averageUnc": "0.8", "trendUnc": "1.2"}, {"date": "1983.12", "average": "1639.4", "trend": "1638.3", "averageUnc": "0.9", "trendUnc": "1.1"}, {"date": "1984.1", "average": "1638.6", "trend": "1639.2", "averageUnc": "1.9", "trendUnc": "1.0"}, {"date": "1984.2", "average": "1638.8", "trend": "1640.1", "averageUnc": "2.0", "trendUnc": "0.9"}, {"date": "1984.3", "average": "1640.7", "trend": "1641.1", "averageUnc": "1.6", "trendUnc": "0.8"}, {"date": "1984.4", "average": "1643.7", "trend": "1642.1", "averageUnc": "1.9", "trendUnc": "0.8"}, {"date": "1984.5", "average": "1642.9", "trend": "1643.1", "averageUnc": "1.1", "trendUnc": "0.7"}, {"date": "1984.6", "average": "1639.6", "trend": "1644.2", "averageUnc": "0.9", "trendUnc": "0.7"}, {"date": "1984.7", "average": "1637.7", "trend": "1645.3", "averageUnc": "1.2", "trendUnc": "0.7"}, {"date": "1984.8", "average": "1641.3", "trend": "1646.4", "averageUnc": "1.6", "trendUnc": "0.6"}, {"date": "1984.9", "average": "1650.4", "trend": "1647.6", "averageUnc": "1.3", "trendUnc": "0.6"}, {"date": "1984.10", "average": "1654.4", "trend": "1648.7", "averageUnc": "1.5", "trendUnc": "0.6"}, {"date": "1984.11", "average": "1653.6", "trend": "1649.8", "averageUnc": "1.2",

Image - Creating the initial json file (for reference)

Creating the desired dataframe - the date column becomes altered?

Answer 1

首先使用convert_dates=False 绕过自动解析并使用dtype={'date': str} 强制将date 列作为原始字符串：

df = pd.read_json('yearly_methane.json', convert_dates=False, dtype={'date': str})

	date	average	trend	averageUnc	trendUnc
0	1983.7	1625.9	1634.6	2.4	1.5
1	1983.8	1628.1	1635.3	2.9	1.4
2	1983.9	1638.4	1636.0	2.2	1.3
3	1983.10	1644.8	1636.7	1.4	1.3

然后使用pd.to_datetime手动将%Y.%m日期转换为datetime对象：

df.date = pd.to_datetime(df.date, format='%Y.%m')

	date	average	trend	averageUnc	trendUnc
0	1983-07-01	1625.9	1634.6	2.4	1.5
1	1983-08-01	1628.1	1635.3	2.9	1.4
2	1983-09-01	1638.4	1636.0	2.2	1.3
3	1983-10-01	1644.8	1636.7	1.4	1.3