The frequencies of the keywords are grouped by class (how-to-do-it, debug-corrective and others).
This means that Java programmers can write a program, load it onto the device, run it, and debug it, with basic programming knowledge of Java, Meike says.
Logging all actions performed in pipelines is important for three reasons: (i) it helps debugging; (ii) it improves repeatability; and (iii) it performs audits in cases where detailed documentation and logging are required by regulatory authorities (such as clinical trials).
Table 16 Results on API elements actually contained in relevant content of the sampled threads, for both APIs and both types of discussion API Discussion type # Elements Relevant for documenting Swing how-to-do-it 115 89.57 debug-corrective 118 50 Android how-to-do-it 144 89.58 debug-corrective 194 43.3.
Table 15 Proportion of occurrence of API element which threads are relevant for documenting such, for both APIs and both types of discussion API Discussion type Relevant for documenting Swing how-to-do-it 94.35 debug-corrective 40.71 Android how-to-do-it 86.82 debug-corrective 23.89.
a, b Swing coverage on threads with how-to-do-it and debug-corrective question, respectively; c, d Android coverage on threads with how-to-do-it and debug-corrective question, respectively.
We adopted their keywords of the how-to-do-it and debug-corrective classes, adding some extra missing keywords.
First, threads are filtered on their respective type of analysis (how-to-do-it or debug-corrective).
Tables 24 and 25 present the numbers of association rules obtained for Swing and for Android, respectively, from both how-to-do-it and debug-corrective coverage analysis.
Also the classifier built by Campos and Maia [15] does not focus only on how-to-do-it and debug-corrective.
In other words, those threads may be potentially useful to document those API elements based on their nature: how-to-do-it or debug-corrective.
