There are usually several possible solutions for most life problems. The same is true for algorithmics, as there are several algorithmic solutions for almost every problem. In the end, we choose the one that has the best quality. The latter is usually estimated by comparing the results of the execution of different algorithms. Usually, these results contain huge amounts of data, so it is often easiest to compare them with the help of graphical representations.
A system for automatic evaluation of algorithms named ALGator was developed at the Faculty of Computer and Information Science in Ljubljana. As the name suggests, it simplifies and automates the algorithm evaluation process. With it, the user can easily and quickly get the results from the execution of algorithms. To graphically compare this data, the existing system only allowed the display of some simple graphs and tables. There was a demand for new displays that would support more complex data analysis, and we tried to meet these demands with this master's thesis.
We first examined the two most commonly used graphical representations, the table and the graph. We gave a general description and presented their examples of good practice and their adaptations for people with limitations. Each version of the table and chart is only suitable for a specific type of data. Therefore, in the thesis we reviewed the data in algorithmics, researched all suitable types of tables and graphs, and adapted the researched versions for representations of the quality of algorithms. We chose the best among them and implemented them in the ALGator system. At the end of the thesis, we presented the ALGator system and the tools used in the implementation of graphical representations and described the implementations.
|
