Analisis Efektifitas Algoritma FAST Menggunakan Metrik Average Percentage Fault Detection dan Waktu Eksekusi Pada Test Case Prioritization

Abstract

It uses regression testing to validate the modification results so as not to cause new errors in features that are already functioning correctly. The implementation of regression testing is generally done by re-executing the test suite used in the previous test. One of the crucial issues in regression testing is determining a strategic approach to reuse existing test suites. Testing with retesting all test cases will result in a long and expensive test. Thus, it is necessary to use a test case selection approach. The research study focuses on analyzing the effectiveness of the FAST algorithm, namely the FAST-pw and FAST-all algorithms, for executing test cases based on priority. The research method uses experiments through running algorithms on test cases that have been implanted with faults in the software with a test case size of 10,000 to 20,000 lines of the program. The effectiveness parameter uses the test case execution time and the percentage of fault detection using the average percentage fault detection (APFD) metric. The results showed that the FAST algorithm, namely the FAST-pw and FAST-all algorithms, had good effectiveness values ​​when applied to TCP testing with small to medium-sized SUTs, namely test cases that tested 1,000 – 20,000 program lines.