Home > Research > Publications & Outputs > A Longitudinal Study of Anti Micro Patterns in ...

Electronic data

  • promise18-14

    Rights statement: © 2018 ACM. This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive Version of Record was published in PROMISE'18 Proceedings of the 14th International Conference on Predictive Models and Data Analytics in Software Engineering http://dx.doi.org/10.1145/3273934.3273945

    Accepted author manuscript, 1.08 MB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

Links

Text available via DOI:

View graph of relations

A Longitudinal Study of Anti Micro Patterns in 113 Versions of Tomcat

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paperpeer-review

Published
Close
NullPointerException

Abstract

Background: Micro patterns represent design decisions in code. They are similar to design patterns and can be detected automatically. These micro structures can be helpful in identifying portions of code which should be improved (anti-micro patterns), or other well-designed parts which need to be preserved. The concepts expressed in these design decisions are defined at class-level; therefore the primary goal is to detect and provide information related to a specific granularity level. Aim: this paper aims to present preliminary results about a longitudinal study performed on anti-micro pattern distributions over 113 versions of Tomcat. Method: we first extracted the micro patterns from the 113 versions of Tomcat, then found the percentage of classes matching each of the six anti-micro pattern considered for this analysis, and studied correlations among the obtained time series after testing for stationarity, randomness and seasonality. Results: results show that the time series are stationary, not random (except for Function Pointer), and that additional studied are needed for studying seasonality. Regarding correlations, only the Pool and Record time series presented a correlation of 0.69, while moderate correlation has been found between Function Pointer and Function Object (0.58) and between Cobol Like and Pool (0.44).

Bibliographic note

© 2018 ACM. This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive Version of Record was published in PROMISE'18 Proceedings of the 14th International Conference on Predictive Models and Data Analytics in Software Engineering http://dx.doi.org/10.1145/3273934.3273945