python - 仅当值不是特定值时按值过滤数据帧

Question

我想按 JSON 中接收的值过滤数据帧，如下所示:

region = request_json['region_value']
ma = request_json['ma_value']
market = request_json['market_value']
subsegment = request_json['subsegment_value']

filtered =df[
    (df['Region'] == region) &
    (df['MA/Segment'] == ma) &
    (df['Market'] == market) &
    (df['Subsegment'] == subsegment)
]

但是，如果region、ma、market或subsegment等于'All'，我不想将其包含在过滤中。

例如，如果region = 'All'，则代码应如下所示:

filtered =df[
        (df['MA/Segment'] == ma) &
        (df['Market'] == market) &
        (df['Subsegment'] == subsegment)
    ]

我想避免多个 if 并寻求更优雅的解决方案。我欢迎任何想法。

Answer 1

您可以使用np.ological_and.reduce；即使您的列名包含空格、特殊字符等，这也会起作用:

conditions = {
    'Region': region, 'MA/Segment': ma, 'Market': market, 'Subsegment': subsegment}

mask = pd.np.logical_and.reduce([
    df[k] == v for k, v in conditions.items() if v != 'All'])
df[mask]

<小时/>

另一个选项是查询，但假设您的列名称符合有效的Python标识符名称，这将起作用。

query = ' and '.join([
    f'{k}=={repr(v)}' for k, v in conditions.items() if v != 'All'])    
df.query(query)