使用 for 循环创建基于多个条件计算百分比的附加列

Question

我有一个包含调查数据的数据框。调查数据评估了多个因素，例如教练、多样性、参与度等。还有其他几个列可以捕获人口统计数据（例如，年龄、部门等）。我想根据包含评级的列添加列。

添加这些列的目的是 a) 提供正面回应的计数，b) 获得正面回应的百分比（正面回应的数量/该因素中的项目数量）和 c) 获得正面回应的百分比因子级别的有利响应（条件是如果任何项目缺少响应，则在因子级别它将为 NULL）

下表显示了仅包含指导项的所需输出。该表应包含其他评分列，并应适用于多样性、领导力、敬业度等因素。

Coach_q1     Coach_q2    Coach_q3      coach_fav_count  coach_fav_perc  coach_agg_perc 
Favourable   Neutral     Favourable    2                66%            66%        
Favourable   Favourable  Fabourable    3                100%           100%
NaN          Favourable  NaN           1                33%            NaN
Favourable   NaN         Favourable    2                66%            NaN         

以下代码用于获取 _favcount 列和 _fav% 列。 ratingcollist 用于仅对具有这些前缀的列应用转换。但是，我无法获得旨在获得整个因素的有利响应百分比的因素水平列 - 仅当针对该特定因素的所有问题都已回答时（即，如果在特定因素中的任何项目中缺少响应因子，那么该因子将产生一个 NaN 值）。

感谢任何形式的帮助，谢谢。

ratingcollist = ['Coach_','Diversity_','Leadership_','Engagement_']


#create a for loop to get all the columns that match the column list keyword
for rat in ratingcollist:
    cols = df.filter(like=rat).columns

#create 2 new columns for each factor, one for count of Favourable responses and one for percentage of Favourable responses
    if len(cols) > 0:
        df[f'{rat.lower()}fav_count'] = (df[cols] == 'Favourable').sum(axis=1)
        df[f'{rat.lower()}fav_perc'] = (df[f'{rat.lower()}fav_count'] / len(cols)) * 100

Answer 1

如果DataFrame.notna 和DataFrame.all 没有丢失所有值，您可以添加掩码进行测试，并且仅针对DataFrame.loc 中的此添加百分比列：

df = pd.DataFrame({'Coach_q1': ['Favourable', 'Favourable', np.nan, 'Favourable'], 'Coach_q2': ['Neutral', 'Favourable', 'Favourable', np.nan], 'Coach_q3': ['Favourable', 'Favourable', np.nan, 'Favourable']})

---

ratingcollist = ['Coach_','Diversity_','Leadership_','Engagement_']


#create a for loop to get all the columns that match the column list keyword
for rat in ratingcollist:
    cols = df.filter(like=rat).columns
    
    mask = df[cols].notna().all(axis=1)

#create 2 new columns for each factor, one for count 
#of Favourable responses and one for percentage of Favourable responses
    if len(cols) > 0:
        df[f'{rat.lower()}fav_count'] = (df[cols] == 'Favourable').sum(axis=1)
        df[f'{rat.lower()}fav_perc'] = (df.loc[mask, f'{rat.lower()}fav_count'] / len(cols)) * 100

print (df)
     Coach_q1    Coach_q2    Coach_q3  coach_fav_count  coach_fav_perc
0  Favourable     Neutral  Favourable                2       66.666667
1  Favourable  Favourable  Favourable                3      100.000000
2         NaN  Favourable         NaN                1             NaN
3  Favourable         NaN  Favourable                2             NaN