This paper is an implementation of Regression Test Generation by Usage Coverage Driven Clustering on User Traces Additional materials can be found on https://github.com/PHILAE-PROJECT. Reach me at frederic.tamagnan (at) gmail.com if you need additional info.
Scanette and Spree are available in ./data
Install packages with
pip -r requirements.txt
cd ./experiments
python3 ucdc_experiments.py
Two abstract classes are available in ./experiments/ucdc/utils. Create your clustering Pipeline and your sampling strategy as classes that implements those abstract class and add them to ucdc_experiments.py
from abc import ABC, abstractmethod
class ClusteringPipeline(ABC):
@abstractmethod
def preprocessor(self,execution_traces_agilkia_format):
pass
@abstractmethod
def fit_predict(self,preprocessed_execution_traces,k):
pass
class SampleHeuristic(ABC):
@abstractmethod
def tests_extraction(self,execution_traces_agilkia_format,cluster_labels):
pass
Plots and json results are in ./experiments/results. Such as