Credit risk data cleaning and variable screening pipeline for pre-loan modeling. Use when working with raw credit data that needs quality assessment, missing value analysis, or variable selection before modeling. it covers data loading and formatting, abnormal period filtering, missing rate calculat

Data Cleaning and Variable Screening Quick Start Complete Process Description The data cleaning pipeline consists of the following 11 steps, each executed independently without deleting the original data: 1. Get Data Load and format raw data 2. Organization Sample Analysis Statistics of sample count and bad sample rate for each organization 3. Separate OOS Data Separate out of sample (OOS) samples from modeling samples 4. Filter Abnormal Months Remove months with insufficient bad sample count or total sample count 5. Calculate Missing Rate Calculate overall and organization level missing rates for each feature 6. Drop High Missing Rate Features Remove features with overall missing rate exceeding threshold 7. Drop Low IV Features Remove features with overall IV too low or IV too low in too many organizations 8. Drop High PSI Features Remove features with unstable PSI 9. Null Importance Denoising Remove noise features using label permutation method 10. Drop High Correlation Features Remove high correlation features based on original gain 11. Export Report Generate Excel report containing details and statistics of all steps Core Functions Function Purpose Module get dataset() Load and format data references.func org analysis() Organization sample analysis references.func missing check() Calculate missing rate references.func drop abnormal ym() Filter abnormal months references.analysis drop highmiss features() Drop high missing rate features references.analysis drop lowiv features() Drop low IV features references.analysis drop highpsi features() Drop high PSI features references.analysis drop highnoise features() Null Importance denoising references.analysis drop highcorr features() Drop high correlation features references.analysis iv distribution by org() IV distribution statistics references.analysis psi distribution by org() PSI distribution statistics references.analysis value ratio distribution by org() Value ratio distribution statistics references.analysis export cleaning report() Export cleaning report references.analysis Parameter Description Data Loading Parameters DATA PATH : Data file path (best are parquet format) DATE COL : Date column name Y COL : Label column name ORG COL : Organization column name KEY COLS : Primary key column name list OOS Organization Configuration OOS ORGS : Out of sample organization list Abnormal Month Filtering Parameters min ym bad sample : Minimum bad sample count per month (default 10) min ym sample : Minimum total sample count per month (default 500) Missing Rate Parameters missing ratio : Overall missing rate threshold (default 0.6) IV Parameters overall iv threshold : Overall IV threshold (default 0.1) org iv threshold : Single organization IV threshold (default 0.1) max org threshold : Maximum tolerated low IV organization count (default 2) PSI Parameters psi threshold : PSI threshold (default 0.1) max months ratio : Maximum unstable month ratio (default 1/3) max orgs : Maximum unstable organization count (default 6) Null Importance Parameters n estimators : Number of trees (default 100) max depth : Maximum tree depth (default 5) gain threshold : Gain difference threshold (default 50) High Correlation Parameters max corr : Correlation threshold (default 0.9) top n keep : Keep top N features by original gain ranking (default 20) Output Report The generated Excel report contains the following sheets: 1. 汇总 Summary information of all steps, including operation results and conditions 2. 机构样本统计 Sample count and bad sample rate for each organization 3. 分离OOS数据 OOS sample and modeling sample counts 4. Step4 异常月份处理 Abnormal months that were removed 5. 缺失率明细 Overall and organization level missing rates for each feature 6. Step5 有值率分布统计 Distribution of features in different value ratio ranges 7. Step6 高缺失率处理 High missing rate features that were removed 8. Step7 IV明细 IV values of each feature in each organization and overall 9. Step7 IV处理 Features that do not meet IV conditions and low IV organizations 10. Step7 IV分布统计 Distribution of features in different IV ranges 11. Step8 PSI明细 PSI values of each feature in each organization each month 12. Step8 PSI处理 Features that do not meet PSI conditions and unstable organizations 13. Step8 PSI分布统计 Distribution of features in different PSI ranges 14. Step9 null importance处理 Noise features that were removed 15. Step10 高相关性剔除 High correlation features that were removed Features Interactive Input : Parameters can be input before each step execution, with default values supported Independent Execution : Each step is executed independently without deleting original data, facilitating comparative analysis Complete Report : Generate complete Excel report containing details, statistics, and distributions Multi process Support : IV and PSI calculations support multi process acceleration Organization level Analysis : Support organization level statistics and modeling/OOS distinction

datanalysis-credit-risk