使用循环在 Python 中创建面板数据

Question

我正在尝试在 Python 中创建面板数据框，例如5 个国家（A、B、C、D、E），每个国家都有 3 年的数据（2000、2001、2002）。

import numpy as np
import pandas as pd

df = {'id': [1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5],
      'country': ['A', 'B', 'C', 'D', 'E', 'A', 'B', 'C', 'D', 'E', 'A', 'B', 'C', 'D', 'E'],
      'year': [2000, 2001, 2002, 2000, 2001, 2002, 2000, 2001, 2002, 2000, 2001, 2002, 2000, 2001, 2002]
        }
df = pd.DataFrame(df)
df

为了将其扩展到更大的数据集，我尝试使用以下代码循环以获得上述结果，但它没有给我所需的数据框。

n_country = 5 # number of countries
n_year = 3 # number of years of data for each country
columns = ("id", "country", "year")
n_rows = n_country*n_year
data = pd.DataFrame(np.empty(shape = (n_rows, 3)), columns = columns)
# set country numbers which will identify each country, create country id ranging from 1 to 5
country_id = range(1, 1 + n_country)
list(country_id)
# create year from 2000 to 2002
year = range(2000, 2000 + n_year)
list(year)
# create dictionary that maps from country id to country name
country_name = dict(zip(country_id, ['A', 'B', 'C', 'D', 'E']))
country_name
# loop starts here
i = 0
for id in country_id:
    for country in ["A", "B", "C", "D", "E"]:
        for year in [2000, 2001, 2002]:
            data.loc[i, "id"] = id
            data.loc[i, "year"] = year
            data.loc[i, "country"] = country_name[id]
            i = +1

生成的数据框不是预期的。

如有网友指出上述循环中的错误，将不胜感激。

谢谢！

Answer 1

我会在 year/countries 上使用产品，然后使用 cat.codes 来标记国家/地区。

from itertools import product
import pandas as pd

start_year = 2000
end_year = 2003

countries = ['A','B','C','D','E']


df = pd.DataFrame(list(product(range(start_year,end_year+1),countries)), columns=['year','country'])
df['id'] = df.country.astype('category').cat.codes+1
print(df)

输出

 year country  id
0   2000       A   1
1   2000       B   2
2   2000       C   3
3   2000       D   4
4   2000       E   5
5   2001       A   1
6   2001       B   2
7   2001       C   3
8   2001       D   4
9   2001       E   5
10  2002       A   1
11  2002       B   2
12  2002       C   3
13  2002       D   4
14  2002       E   5
15  2003       A   1
16  2003       B   2
17  2003       C   3
18  2003       D   4
19  2003       E   5

至于您当前的循环，您可能需要 zip id 和 country，以便每年循环重复使用它们，并且需要 i+=1 而不是 i=+1

n_country = 5 # number of countries
n_year = 3 # number of years of data for each country
columns = ("id", "country", "year")
n_rows = n_country*n_year
data = pd.DataFrame(np.empty(shape = (n_rows, 3)), columns = columns)
# set country numbers which will identify each country, create country id ranging from 1 to 5
country_id = range(1, 1 + n_country)
list(country_id)
# create year from 2000 to 2002
year = range(2000, 2000 + n_year)
list(year)
# create dictionary that maps from country id to country name
country_name = dict(zip(country_id, ['A', 'B', 'C', 'D', 'E']))
country_name
# loop starts here
i = 0
for c_id,country in zip(country_id,["A", "B", "C", "D", "E"]):
    print(c_id, country)
    for year in [2000, 2001, 2002]:
        data.loc[i, "id"] = c_id
        data.loc[i, "year"] = year
        data.loc[i, "country"] = country
        i +=1

使用循环在 Python 中创建面板数据

Create panel data in Python using loop

python

loops

panel