根据两个日期列之间的差异过滤 Pandas 数据框的最有效方法是什么?
例如,基于以下数据帧:
CADASTRO RESPOSTA EVAL
0 2021-06-01 2021-06-13 y
1 2021-06-01 2021-06-13 y
2 2021-06-01 2021-06-18 y
3 2021-06-01 2021-06-09 n
4 2021-06-01 2021-06-20 n
5 2021-06-01 2021-06-20 n
如何过滤它以仅包含列 RESPOSTA
和 CADASTRO
之间的差异小于 15 天的记录?我尝试了以下方法,但没有成功:
import datetime
filtered_df = df[(df.RESPOSTA - df.CADASTRO).days < 15]
所需的输出是:
CADASTRO RESPOSTA EVAL
0 2021-06-01 2021-06-13 y
1 2021-06-01 2021-06-13 y
3 2021-06-01 2021-06-09 n
答案 0 :(得分:3)
通过日期时间访问器 dt
# Ensure DateTime
df['CADASTRO'] = pd.to_datetime(df['CADASTRO'])
df['RESPOSTA'] = pd.to_datetime(df['RESPOSTA'])
# Access Days through dt.days
filtered_df = df[(df.RESPOSTA - df.CADASTRO).dt.days < 15]
filtered_df
:
CADASTRO RESPOSTA EVAL
0 2021-06-01 2021-06-13 y
1 2021-06-01 2021-06-13 y
3 2021-06-01 2021-06-09 n
答案 1 :(得分:1)
使用timedelta
比较天差
from datetime import timedelta
df = pd.DataFrame({
'CADASTRO': ['2021-06-01', '2021-06-01', '2021-06-01', '2021-06-01', '2021-06-01', '2021-06-01'],
'RESPOSTA': ['2021-06-13', '2021-06-13', '2021-06-18', '2021-06-09', '2021-06-20', '2021-06-20'],
'EVAL': ['y', 'y', 'y', 'n', 'n', 'n']
})
df['CADASTRO'] = pd.to_datetime(df['CADASTRO'])
df['RESPOSTA'] = pd.to_datetime(df['RESPOSTA'])
df['temp'] = df['RESPOSTA'] - df['CADASTRO']
df['temp'] = df['temp'].apply(lambda x: 0 if x < timedelta(days=15) else 1)
filtered_df = df.drop(df[df['temp']==0].index).drop(columns=['temp'])
输出filtered_df