我有一个看起来像这样的数据集-
yyyy month tmax tmin
0 1908 January 5.0 -1.4
1 1908 February 7.3 1.9
2 1908 March 6.2 0.3
3 1908 April 7.4 2.1
4 1908 May 16.5 7.7
5 1908 June 17.7 8.7
6 1908 July 20.1 11.0
7 1908 August 17.5 9.7
8 1908 September 16.3 8.4
9 1908 October 14.6 8.0
10 1908 November 9.6 3.4
11 1908 December 5.8 -0.3
12 1909 January 5.0 0.1
13 1909 February 5.5 -0.3
14 1909 March 5.6 -0.3
15 1909 April 12.2 3.3
16 1909 May 14.7 4.8
17 1909 June 15.0 7.5
18 1909 July 17.3 10.8
19 1909 August 18.8 10.7
20 1909 September 14.5 8.1
21 1909 October 12.9 6.9
22 1909 November 7.5 1.7
23 1909 December 5.3 0.4
24 1910 January 5.2 -0.5
...
它具有四个变量-
yyyy，month，tmax(最高温度)和tmin我想在预测时将月份列用作变量，因此要将其转换为二进制编码版本。本质上，我想在名为
January的数据集中添加十二个变量，直到December为止，如果特定行的月份为“January”，则应将January列标记为1，其余11列新添加的列应为0。我查看了数据透视表，但这对我的事业没有帮助。关于如何以简单优雅的方式执行此操作的任何想法？
最佳答案
我认为您需要 get_dummies :
df = pd.get_dummies(df['month'])
如果需要在原始列中添加新列并删除
month，请使用 join 和 pop :df2 = df.join(pd.get_dummies(df.pop('month')))
print (df2.head())
yyyy tmax tmin April August December February January July June \
0 1908 5.0 -1.4 0 0 0 0 1 0 0
1 1908 7.3 1.9 0 0 0 1 0 0 0
2 1908 6.2 0.3 0 0 0 0 0 0 0
3 1908 7.4 2.1 1 0 0 0 0 0 0
4 1908 16.5 7.7 0 0 0 0 0 0 0
March May November October September
0 0 0 0 0 0
1 0 0 0 0 0
2 1 0 0 0 0
3 0 0 0 0 0
4 0 1 0 0 0
如果不需要删除列
month:df2 = df.join(pd.get_dummies(df['month']))
print (df2.head())
yyyy month tmax tmin April August December February January \
0 1908 January 5.0 -1.4 0 0 0 0 1
1 1908 February 7.3 1.9 0 0 0 1 0
2 1908 March 6.2 0.3 0 0 0 0 0
3 1908 April 7.4 2.1 1 0 0 0 0
4 1908 May 16.5 7.7 0 0 0 0 0
July June March May November October September
0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0
2 0 0 1 0 0 0 0
3 0 0 0 0 0 0 0
4 0 0 0 1 0 0 0
如果需要排序列，则有更多可能的解决方案-使用
reindex 或 reindex_axis :months = ['January', 'February', 'March','April' ,'May', 'June', 'July', 'August', 'September','October', 'November','December']
df1 = pd.get_dummies(df['month']).reindex_axis(months, 1)
print (df1.head())
January February March April May June July August September \
0 1 0 0 0 0 0 0 0 0
1 0 1 0 0 0 0 0 0 0
2 0 0 1 0 0 0 0 0 0
3 0 0 0 1 0 0 0 0 0
4 0 0 0 0 1 0 0 0 0
October November December
0 0 0 0
1 0 0 0
2 0 0 0
3 0 0 0
4 0 0 0
df1 = pd.get_dummies(df['month']).reindex(columns=months)
print (df1.head())
January February March April May June July August September \
0 1 0 0 0 0 0 0 0 0
1 0 1 0 0 0 0 0 0 0
2 0 0 1 0 0 0 0 0 0
3 0 0 0 1 0 0 0 0 0
4 0 0 0 0 1 0 0 0 0
October November December
0 0 0 0
1 0 0 0
2 0 0 0
3 0 0 0
4 0 0 0
或将
month列转换为ordered categorical:df1 = pd.get_dummies(df['month'].astype('category', categories=months, ordered=True))
print (df1.head())
January February March April May June July August September \
0 1 0 0 0 0 0 0 0 0
1 0 1 0 0 0 0 0 0 0
2 0 0 1 0 0 0 0 0 0
3 0 0 0 1 0 0 0 0 0
4 0 0 0 0 1 0 0 0 0
October November December
0 0 0 0
1 0 0 0
2 0 0 0
3 0 0 0
4 0 0 0
关于python - Pandas -将分类列转换为二进制编码形式，我们在Stack Overflow上找到一个类似的问题：https://stackoverflow.com/questions/45416212/