熊猫从 CSV 和 groupby 每月总工作日读取日期

Question

data.CSV

ID Activity Month   Activity Date

0   04/2019     04-01-2019

1   05/2019     05-13-2019

2   05/2019     05-25-2019

3   06/2019     06-10-2019

4   06/2019     06-19-2019

5   07/2019     07-15-2019

6   07/2019     07-18-2019

7   07/2019     07-29-2019

8   08/2019     06-03-2019

9   08/2019     06-15-2019

10  08/2019     06-20-2019

我的计划

读取 csv：

df = pd.read_csv('data.CSV')

转换为日期时间：

df['活动日期'] = pd.to_datetime(df['活动日期'], dayfirst=True)

按活动月份列分组：

grouped = df.groupby(['活动月份'])['活动日期'].count()

打印（分组）

Activity Month
04/2019    15532
05/2019    13924
06/2019    12822
07/2019    14067
08/2019    10939
Name: Activity Date, dtype: int64

对日期进行分组时，执行工作日计算：

这部分我不知道该怎么做。已经迷路了

我用来计算工作日的代码

import calendar
import datetime

x = datetime.date(2019, 4, 1)
cal = calendar.Calendar()
working_days = len([x for x in cal.itermonthdays2(x.year, x.month) if x[0] !=0 and x[1] < 5])
print ("Total business days for month (" + str(x.month) +  ") is " + str(working_days) + " days")

我想要的输出

Total business days for month (4) is 22 days
Total business days for month (5) is 23 days
Total business days for month (6) is 20 days
Total business days for month (7) is 23 days
Total business days for month (8) is 22 days

Answer 1

我并不完全清楚这里的问题陈述，但是如果您想计算每个 Activity Month 的工作日数，您可以将计算包装在一个方法中，然后将该方法应用于您的 Activity Month列（lambda 表达式基本上是针对指定列的每一行的 for 循环操作）。

grouped = df.groupby(['Activity Month'])['Activity Date'].count().reset_index()

def get_business_days(x):
    x = datetime.date(int(x.split('/')[1]), int(x.split('/')[0]), 1)
    cal = calendar.Calendar()
    working_days = len([x for x in cal.itermonthdays2(x.year, x.month) if x[0] !=0 and x[1] < 5])
    return ("Total business days for month (" + str(x.month) +  ") is " + str(working_days) + " days")

grouped['Activity Month'].apply(get_business_days)

输出是具有您的文本输出的系列。

0    Total business days for month (4) is 22 days
1    Total business days for month (5) is 23 days
2    Total business days for month (6) is 20 days
3    Total business days for month (7) is 23 days
4    Total business days for month (8) is 22 days

但是，在每个单元格中存储重复信息是个坏主意。最好简单地返回 working_days 而不是将其嵌入字符串中。