https://github.com/fengxiaoxiao-001/data_preprocessing

提供处理缺失值，处理异常值，处理特征工程以及多种数据绘图功能;适合大型数据，以及配备处理超多不同数据类型分布的方法
https://github.com/fengxiaoxiao-001/data_preprocessing

data-science data-visualization processing

Last synced: 3 days ago
JSON representation

提供处理缺失值，处理异常值，处理特征工程以及多种数据绘图功能;适合大型数据，以及配备处理超多不同数据类型分布的方法

• Host: GitHub
• URL: https://github.com/fengxiaoxiao-001/data_preprocessing
• Owner: Fengxiaoxiao-001
• License: apache-2.0
• Created: 2025-05-26T14:54:26.000Z (11 months ago)
• Default Branch: main
• Last Pushed: 2025-06-02T12:36:52.000Z (11 months ago)
• Last Synced: 2025-06-17T18:03:49.527Z (11 months ago)
• Topics: data-science, data-visualization, processing
• Language: Python
• Homepage:
• Size: 82 KB
• Stars: 1
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Data_preprocessing-数据预处理

两个方便的类，分别用于处理数据的缺失值和异常值  

DataLack: 负责处理缺失值  

Outliers:负责处理异常值  

Two convenient classes for handling missing and outliers in your data  

DataLack: Responsible for handling missing values  

Outliers: Responsible for handling outliers  

# JudgeMethod流程图

graph TD  

    A[judge_distribution] --> B{数值型?}  

    B -->|否| C[返回 categorical]  

    B -->|是| D[计算nunique/total]  

    D --> E{离散型条件?}  

    E -->|是| F[_judge_discrete]  

    E -->|否| G[_judge_continuous]  

    G --> H[计算偏度/峰度]  

    H --> I[分层抽样]  

    I --> J[正态性检验]  

    J --> K{是否正态?}  

    K -->|是| L[返回normal]  

    K -->|否| M{是否复杂分布?}  

    M -->|是| N[拟合分布+形状检测]  

    M -->|否| O[返回矩量判断结果]  

    F --> P{唯一值≤2?}  

    P -->|是| Q[二项分布检验]  

    P -->|否| R[泊松/负二项检验]  

    Q & R --> S[返回离散型分布]