根据两个不同列中的各自值在 DataFrame 中创建新列答案

【问题标题】：Create new column in DataFrame based on respective values in two different columns根据两个不同列中的各自值在 DataFrame 中创建新列
【发布时间】：2020-11-28 14:44:40
【问题描述】：

我有以下数据框：

cars = {'Brand': ['Honda Civic','Toyota Corolla','Ford Focus','No Brand'],
        'Country': ['Japan','No Country','United States','Germany']
        }

df = pd.DataFrame(cars, columns = ['Brand', 'Country'])
df.head(4)

            Brand        Country
0     Honda Civic          Japan
1  Toyota Corolla     No Country
2      Ford Focus  United States
3        No Brand        Germany

希望在数据框中创建一个新列，该列将根据“品牌”和“国家”列的值进行组合。如果 Brand 列中有“No Brand”值，则 Desc 列仅采用 Country 列中的值。如果 Country 列中有“No Country”值，则 Desc 列仅采用 Brand 列中的值。期望的输出：

            Brand        Country    Desc
0     Honda Civic          Japan    Honda Civic Japan
1  Toyota Corolla     No Country    Toyota Corolla
2      Ford Focus  United States    Ford Focus United States
3        No Brand        Germany    Germany

如果它检查一列中的字符串，我可以这样做，但不确定如何处理两列。现在我只能在我想要的条件下检查布尔值。

df['Desc'] = df['Brand'].str.contains("No Brand") | df['Country'].str.contains("No Country")

            Brand        Country    Desc
0     Honda Civic          Japan    False
1  Toyota Corolla     No Country    True
2      Ford Focus  United States    False
3        No Brand        Germany    True

我读到不建议迭代数据帧并避免这样做。

【问题讨论】：

标签： python pandas numpy dataframe

【解决方案1】：

def get_desc(brand, country):
    return (brand if brand != 'No Brand' else '') +\
           (' ' + country if country != 'No Country' else '')


df['Desc'] = df['Brand'].combine(df['Country'], get_desc)

print(df.head(4))

输出：

            Brand        Country                      Desc
0     Honda Civic          Japan         Honda Civic Japan
1  Toyota Corolla     No Country            Toyota Corolla
2      Ford Focus  United States  Ford Focus United States
3        No Brand        Germany                   Germany

【讨论】：

【解决方案2】：

让我们concat 这两列然后使用str.replace 将No Brand 和No Country 值替换为空字符串：

df['Desc'] = (df['Brand'] + ' ' + df['Country']).str.replace(r'No Brand\s*|\s*No Country', '')

结果：

            Brand        Country                      Desc
0     Honda Civic          Japan         Honda Civic Japan
1  Toyota Corolla     No Country            Toyota Corolla
2      Ford Focus  United States  Ford Focus United States
3        No Brand        Germany                   Germany

【讨论】：

【解决方案3】：

In [2]: cars = {'Brand': ['Honda Civic','Toyota Corolla','Ford Focus','No Brand'],
   ...:         'Country': ['Japan','No Country','United States','Germany']
   ...:         }
   ...: 
   ...: df = pd.DataFrame(cars, columns = ['Brand', 'Country'])
   ...: df
Out[2]: 
            Brand        Country
0     Honda Civic          Japan
1  Toyota Corolla     No Country
2      Ford Focus  United States
3        No Brand        Germany

In [3]: df['New_col'] = (df.Brand + " " + df.Country).str.replace("No Brand", "").str.replace("No Country", "").str.strip()

In [4]: df
Out[4]: 
            Brand        Country                   New_col
0     Honda Civic          Japan         Honda Civic Japan
1  Toyota Corolla     No Country            Toyota Corolla
2      Ford Focus  United States  Ford Focus United States
3        No Brand        Germany                   Germany

【讨论】：