如何使用Pandas的DataFrame统计一个月内每天的病例数？

Question

I would like to count the number of daily cases within a certain month as follows:

import pandas as pd

d1 = pd.DataFrame({'ID': ["A", "A", "A", "B", "B", "C", "C", "C", "C", "D", "D", "D"],
                   "date": ["2010-12-30", "2010-02-27", "2010-02-26", "2012-01-01", "2012-01-03",
                            "2011-01-01", "2011-01-02", "2011-01-08", "2014-02-21", "2010-08-31", "2010-08-30", "2010-09-01"]})

and the final outcome would be like this:

  ID year_month  count
0  A    2010-02      2
1  A    2010-12      1
2  B    2012-01      2
3  C    2011-01      3
4  C    2014-02      1
5  D    2010-08      2
6  D    2010-09      1

Do you have any ideas about how to produce the DataFrame like the above? I used groupby and apply functions but could not produce like that. Thanks in advance!

Answer 1

Use Series.dt.to_period for month periods and count by GroupBy.size :

#convert to datetimes if necessary
#d1['date'] = pd.to_datetime(d1['date'])

df = (d1.groupby(['ID', d1['date'].dt.to_period('m').rename('year_month')])
        .size()
        .reset_index(name='count'))
print (df)
  ID year_month  count
0  A    2010-02      2
1  A    2010-12      1
2  B    2012-01      2
3  C    2011-01      3
4  C    2014-02      1
5  D    2010-08      2
6  D    2010-09      1

Another idea with Series.dt.strftime :

#convert to datetimes if necessary
#d1['date'] = pd.to_datetime(d1['date'])

df = (d1.groupby(['ID', d1['date'].dt.strftime('%Y-%m').rename('year_month')])
        .size()
        .reset_index(name='count'))
print (df)
  ID year_month  count
0  A    2010-02      2
1  A    2010-12      1
2  B    2012-01      2
3  C    2011-01      3
4  C    2014-02      1
5  D    2010-08      2
6  D    2010-09      1

If no datetimes, but strings:

df = (d1.groupby(['ID', d1['date'].str[:7].rename('year_month')])
        .size()
        .reset_index(name='count'))
print (df)
  ID year_month  count
0  A    2010-02      2
1  A    2010-12      1
2  B    2012-01      2
3  C    2011-01      3
4  C    2014-02      1
5  D    2010-08      2
6  D    2010-09      1

Answer 2

Using apply and groupby should work:

import pandas as pd

d1 = pd.DataFrame({'ID': ["A", "A", "A", "B", "B", "C", "C", "C", "C", "D", "D", "D"],
                   "date": ["2010-12-30", "2010-02-27", "2010-02-26", "2012-01-01", "2012-01-03",
                            "2011-01-01", "2011-01-02", "2011-01-08", "2014-02-21", "2010-08-31", "2010-08-30", "2010-09-01"]})

d1["month_year"] = d1.apply(lambda row: row["date"][:7])
month_year = d1.groupby("month_year").size().reset_index(name="count")

print(month_year)

This will result:

  month_year  count
0    2010-02      2
1    2010-08      2
2    2010-09      1
3    2010-12      1
4    2011-01      3
5    2012-01      2
6    2014-02      1

You will probably want to change the apply lambda to handle the date more carefully.