使用类时出现 NameError [关闭]答案

【问题标题】：NameError when using a class [closed]使用类时出现 NameError [关闭]
【发布时间】：2021-08-17 13:20:59
【问题描述】：

我想为目标变量连续的数据集设置修改后的 WOE 和 IV 摘要（受此网站启发：https://www.listendata.com/2019/08/WOE-IV-Continuous-Dependent.html#comment-form）。代码是：

class IV_Calc:
    def __init__(self, df, feature, target):
        self.feature = feature
        self.target = target
        self.data_head = df.head()

    def count_values(self):
        data = pd.DataFrame()
        data['Count'] = df[self.feature].value_counts()               
        
        data['Sum Y'] = df.groupby([self.feature])[self.target].sum()
        
        data = data.sort_values(by=["Count"], ascending=False)
    
    
        return data

    def distribution(self):
        data = self.count_values()
        data['% Observations'] = data['Count'] / data
        data['% Y'] = data['Sum Y'] / data['Sum Y'].sum()
        
        return data.iloc[:,-2:]
  
    def woe(self):
        data = self.distribution()
        data['WOE'] = np.log(data['% Y'] / data['% Observations'])
        data.replace({"WOE": {np.inf: 0, -np.inf: 0}})  # If no instances are bad, this will replace values of infinity with 0
        data = data.sort_values(by=["WOE"], ascending=False)
        return data.iloc[:,-1]
  
  
    def IV_per_cat(self):
        data = self.distribution()
        data['WOE'] = self.woe()
        data["IV"] = data["WOE"]*(data['% Y'] - data['% Observations'])
        data = data.sort_values(by=["IV"], ascending=False)
        return data.iloc[:,-1]
  
    def full_summary(self):
        data = self.count_values()
        data['% Observations'] = data['Count'] / data
        data['% Y'] = data['Sum Y'] / data['Sum Y'].sum()
        data['WOE'] = self.woe()
        data["IV"] = self.IV_per_cat()
        data = data.sort_values(by=["Count"], ascending=False)
        return data

    def final_assessment(self):
        data = self.full_summary()
        iv = data["IV"].sum() # final IV value
        if iv < .02:
            print("The variable " + self.feature + " is not predictive with an IV of: {}".format(iv))
        elif iv < .1:
            print("The variable " + self.feature + " is weakly predictive with an IV of:{}".format(iv))
        elif iv < .3:
            print("The variable " + self.feature + " is moderately predictive with an IV of:{}".format(iv))
        else :
            print("The variable " + self.feature + " is highly predictive with an IV of: {}".format(iv))
        return iv

跑步

woe_test = IV_Calc(df=train, feature="host_response_time", target="price")

woe_test.full_summary()

虽然我在__init__ 部分定义了df，但第一个函数出现错误“名称'df' 未定义”。我在监督什么？我确信这与我使用类有关，但我在面向对象编程方面经验不足，无法找出它是什么。

【问题讨论】：

标签： python function class

【解决方案1】：

您不会将 df 分配给实例变量。

你需要这样做

class IV_Calc:
    def __init__(self, df, feature, target):
        self.feature = feature
        self.target = target
        self.df = df  # <-- this right here

然后在类的方法中使用self.df 而不是df。

【讨论】：