在带有日期的几列上具有多个条件的子集 pandas 数据框 [英] Subset pandas data frame with multiple condition on several columns with date

问题描述

我有一个数据框，如下所示:

I have a data frame as follows:

  slot_id class        day   base_date
0        1     A     Monday  2019-01-21
1        2     B    Tuesday  2019-01-22
2        3     C  Wednesday  2019-01-23
3        4     C  Wednesday  2019-01-23
4        5     C   Thursday  2019-01-24

具有以下信息:

example.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8 entries, 0 to 7
Data columns (total 4 columns):
slot_id      8 non-null int64
class        8 non-null object
day          8 non-null object
base_date    8 non-null object
dtypes: int64(1), object(3)
memory usage: 200.0+ bytes

我想获得具有最小base_date的class == "C"的行.我已经尝试了许多min组合，但都没有成功，就像这样:

I would like to get the row with class == "C" with the minimum base_date. I have tried many combinations of min without success, like this one:

example[(example['class']=="C") & min(example['base_date'])]

上面的代码有什么问题?如何获得满足以上条件的完整行? 完整的example数据帧:

What's wrong with the code above? How could I get a full row meeting the condition above? Full example data frame:

{'slot_id': {0: 1, 1: 2, 2: 3, 3: 4, 4: 5, 5: 6, 6: 7, 7: 8}, 'class': {0: 'A', 1: 'B', 2: 'C', 3: 'C', 4: 'C', 5: 'C', 6: 'D', 7: 'E'}, 'day': {0: 'Monday', 1: 'Tuesday', 2: 'Wednesday', 3: 'Wednesday', 4: 'Thursday', 5: 'Tuesday', 6: 'Thursday', 7: 'Saturday'}, 'base_date': {0: datetime.date(2019, 1, 21), 1: datetime.date(2019, 1, 22), 2: datetime.date(2019, 1, 23), 3: datetime.date(2019, 1, 23), 4: datetime.date(2019, 1, 24), 5: datetime.date(2019, 1, 22), 6: datetime.date(2019, 1, 24), 7: datetime.date(2019, 1, 26)}}
example['base_date'] = pd.to_datetime(example['base_date'].astype(str), format='%d%m%Y')
example['base_date'] = example['base_date'].dt.date

推荐答案

您可以使用idxmin:

pd.to_datetime(df.loc[df['class'] == 'C', 'base_date']).idxmin()
# 2

df.iloc[pd.to_datetime(df.loc[df['class'] == 'C', 'base_date']).idxmin()]

slot_id               3
class                 C
day           Wednesday
base_date    2019-01-23
Name: 2, dtype: object

---

如果您需要重复执行此操作，更好的解决方案是将"base_date"预先转换为datetime类型:

---

If you need to do this repeatedly, a better solution is to pre-convert "base_date" to datetime type:

df['base_date'] = pd.to_datetime(df['base_date'], errors='coerce')
df.iloc[df.loc[df['class'] == 'C', 'base_date'].idxmin()]

slot_id                        3
class                          C
day                    Wednesday
base_date    2019-01-23 00:00:00
Name: 2, dtype: object

这篇关于在带有日期的几列上具有多个条件的子集 pandas 数据框的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！