# METTLE

**Repository Path**: xiaoyuanxie/mettle

## Basic Information

- **Project Name**: METTLE
- **Description**: METTLE：基于蜕变测试与失效模式的系统度量与选择工具
- **Primary Language**: Unknown
- **License**: GPL-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-04-19
- **Last Updated**: 2021-04-20

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# METTLE: a METmorphic Testing approach to assessing and validating unsupervised machine LEarning systems

## File structure (main)

- [cluster](mettle/src/main/java/org/whu/mettle/cluster): clustering implementation
- [mr](mettle/src/main/java/org/whu/mettle/mr): MR implmentation
- [util](mettle/src/main/java/org/whu/mettle/util): utils
- [evaluate](mettle/src/main/java/org/whu/mettle/evaluate): clustering assessment (check violation & compute score)
- [main](mettle/src/main/java/org/whu/mettle/main): executive module
- [testdata](mettle/src/main/resources/testdata): source dataset to be executed
- [lib](mettle/lib): auxiliary python script

## To run

in `main` directory:

- identify test cases (e.g., system under test, source dataset to be executed) in **MainTest.csv**
- running **MainTest.java** (identify MRs and relevant weights in terms of `MR preference` before running), MR compliance will be automatically generated as `evaluate.csv`
- identify weights in terms of `MR serverity` in `evaluate.csv` (after your further assessment), running **Evaluate.java** to acquire final ranking of investigated clustering systems


## Note

Python environments is addtionally required to support scientific computation (e.g., numpy, pandas, scikit-learn)