pandas使用cut函数基于分位数进行连续值分箱（手动计算分位数）处理后出现NaN值原因及解决

Posted 2021-10-31 Data+Science+Insight

pandas使用cut函数基于分位数进行连续值分箱（手动计算分位数）处理后出现NaN值原因及解决

目录

pandas使用cut函数基于分位数进行连续值分箱（手动计算分位数）处理后出现NaN值原因及解决

问题：

解决：

---

['bad', 'good', 'medium', 'medium', 'good', ..., NaN, NaN, 'medium', 'good', 'awesome']
Length: 13
Categories (4, object): ['bad' < 'medium' < 'good' < 'awesome']

问题：

test = np.array([1, 7, 5, 4, 6, 3,20,10,0,0,4,9,30])

percentiles = np.percentile(test, [0, 25, 50, 75,100])

percentiles = list(percentiles)

percentiles


pd.cut(test,bins = percentiles, labels=["bad", "medium", "good", "awesome"])

[0.0, 3.0, 5.0, 9.0, 30.0]
 

['bad', 'good', 'medium', 'medium', 'good', ..., NaN, NaN, 'medium', 'good', 'awesome']
Length: 13
Categories (4, object): ['bad' < 'medium' < 'good' < 'awesome'] 

解决：

#include_lowest = True

#include_lowest = True

test = np.array([1, 7, 5, 4, 6, 3,20,10,0,0,4,9,30])

percentiles = np.percentile(test, [0, 25, 50, 75,100])

percentiles = list(percentiles)

percentiles


pd.cut(test,bins = percentiles, labels=["bad", "medium", "good", "awesome"],include_lowest = True)

[0.0, 3.0, 5.0, 9.0, 30.0]

['bad', 'good', 'medium', 'medium', 'good', ..., 'bad', 'bad', 'medium', 'good', 'awesome']
Length: 13
Categories (4, object): ['bad' < 'medium' < 'good' < 'awesome']

参考：python

参考：pandas