No abstract

There has been considerable discussion on the possible impacts of open source software development practices, especially in regard to the quality of the resulting software product. Recent studies have shown that analyzing data from source code repositories is an efficient way to gather information about project characteristics and programmers, showing that OSS projects are very heterogeneous in their team structures and software processes. However, one problem is that the resulting process metrics measuring attributes of the development process and of the development environment do not give any hints about the quality, complexity, or structure of the resulting software. Therefore, we expanded the analysis by calculating several product metrics, most of them specifically tailored to object-oriented software. We then analyzed the relationship between these product metrics and process metrics derived from a CVS repository. The aim was to establish whether different variants of open source development processes have a significant impact on the resulting software products. In particular we analyzed the impact on quality and design associated with the numbers of contributors and the amount of their work, using the GINI coefficient as a measure of inequality within the developer group.

No abstract

Black-box models such as linear regression have proven to be helpful in ongoing building commissioning in many ways. The aim of this work is to improve linear models with change point for fault detection in buildings. Building simulations revealed poor performance of them (R2 < 0.7) for some low energy buildings. The regression models (RMs) can be considerably improved by introducing the rate of change of the indoor air temperature (Tind) as an independent variable. Thus, R2 values were raised by up to 0.5 (e.g. from 0.2 to 0.7, example with the lowest R2). A new training and application process for the RMs revealed further improvements by using a hierarchical agglomerative clustering algorithm to determine different day-types as additional (categorical) variables in the RM. The application of these improved RMs for outlier detection is demonstrated in three buildings

No abstract