Fault potential has some relation to many measurable

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: to find? The span of files increases over time (age) • • Tuesday, September 10, 13 • • Modularity breaks over time Effort has some relations to many measurable variables. Fault potential has some relation to many measurable variables. Hypotheses • What the authors are trying or expecting to find? 1. 2. 3. 4. • Tuesday, September 10, 13 The span of changes increases over time Breakdown of modularity increases over time Fault potential, the likelihood of changes to induce faults has some relations to ... Efforts has some relationship too Usually *good* empirical study paper either finds surprising empirical evidence that contradicts conventional wisdom or provides thorough empirical evidence that validates well known hypotheses. Study Approach • • Data selection • Study method that finds *relationships* among the measurement variables Tuesday, September 10, 13 Selection of measurement variables (so called independent variables) Study Approach: (1) Data Selection • Rich data set • This system evolved by following a well-defined process. Structured, manually labeled data Easy to group related changes Telephone switching system 2 • • • • Tuesday, September 10, 13 IEEE TRANSACTIONS ON SOFTWARE ability to maintain ality or fixing faul called ªmaintenan ªreengineering.º 100 million LOC 5000 modules 50 major subsystems in C and C+ Fig. 1. Changes to the code (bold boxes) and associated data fields. Section 5.3, we show that the distribution of faults is explained by the distribution of large, recent changes to the software), and 4) prediction of effort quantifies the impact of decay required to make a change using code decay indices that encapsulate characteristics of changes (Section 5.4). 1.1 Related Work 2.1 The Change For the system we a well-defined proc credit card billing) by which the syste Changes that im sent to the develop Requests (IMRs); i requires hundreds the IMR distributes implementing para unavailable files. Each IMR gene (MRs), which conta be done to each mo an associated MR problem.) To perfo MR, makes the req whether the chan Study Approach: (2) Measure Independent Variables • • c denotes changes (mostly a MR) Variables • • • • • • Tuesday, September 10, 13 DELTAS(c) = # of deltas associated with c ADD(c) = # of lines added by c DEL(c) = # of lines deleted by c DATE(c) = the date on which c is completed INT(c) = the interval of c (calendar time required to implement c) DEV(c) = number of developers implementing c Inexperienced developers can be either new to programming or new to particular code. They increase the risk of decay because of lack of knowledge, a lack of understanding of the system architecture, and (for those early in their careers) potential for lower or less-developed skills. ODE denotes files, and m denotes modules. None of these objects P are subscripted, so that (for example) c denotes a sum over all changes. For a change c and software unit m, c e> m means that ªc touches m:º some part of m is changed by c. Also 1fAg, the indicator of the event A, is equal to one if A occurs and zero otherwise. In addition, several of the CDIs (all computable directly f r o m t h e ve r s i o n m a n a g e m e n t d a t a b a s e ) d e p e n d o n characteristics of changes: Study Approach: (2) Measure Independent Variables DECAY INDICES software engineering literature there is a rich of studies involving software measurement and rement theory (see [17]). Our code decay indices in this tradition, by being both quantified and ble in the version management data base. Pursuing the al metaphor, CDIs may be interpreted as quantified oms, quantified risk factors, or prognoses, which edictors of the responses (cost, interval, quality). arily, prognoses are functions of quantified sympnd risk factors. order to define actionable priorities to remediate indices must encapsulate developer knowledge and inate over both time and location in the software. everal of the indices can be visualized in compelling as we illustrate in Sections 4.2.4 and 5.2. • eneral Considerations ADD…c† ˆ number of lines added by c DEL…c† ˆ numb...
View Full Document

Ask a homework question - tutors are online