在不丢失数据框中已有值的情况下填充数据框中的缺失值

Question

我有缺失值的数据框：

import pandas as pd
data = {'Brand':['residential','unclassified','tertiary','residential','unclassified','primary','residential'],
    'Price': [22000,25000,27000,"NA","NA",10000,"NA"]
    }

df = pd.DataFrame(data, columns = ['Brand', 'Price'])

print (df)

导致此数据框：

             Brand  Price
    0   residential  22000
    1  unclassified  25000
    2      tertiary  27000
    3   residential     NA
    4  unclassified     NA
    5       primary  10000
    6   residential     NA

我想用固定值（住宅=1000，未分类=2000）在价格列中填写住宅和未分类的缺失值，但是我不想丢失住宅或未分类价格列中已经存在的任何值，所以输出应该是这样的：

        Brand  Price
    0   residential  22000
    1  unclassified  25000
    2      tertiary  27000
    3   residential   1000
    4  unclassified   2000
    5       primary  10000
    6   residential   1000

完成这项工作的最简单方法是什么

Answer 1

我们可以map用fillna, PS: 你需要确保在你的 df, NA 是 NaN

df.Price.fillna(df.Brand.map({'residential':1000,'unclassified':2000}),inplace=True)
df
          Brand    Price
0   residential  22000.0
1  unclassified  25000.0
2      tertiary  27000.0
3   residential   1000.0
4  unclassified   2000.0
5       primary  10000.0
6   residential   1000.0