Abstract. We survey several planar graph generators that were selected according to availability, theoretical interest, easiness of implementation and efficiency. We experimentally study graph properties that allow for a basic classification of the generators. This analysis is extended by means of advanced algorithmical behavior on the generated graphs, in particular kernelization of fixed-parameter tractable problems. We will see the major influence of instance selection on algorithmic behavior. This selection has been disregarded in several publications, which deduce general results from non-representative data sets. Altogether, this study helps experimenters to carefully select sets of planar graphs that allow for a meaningful interpretation of their results.

No abstract

No abstract

No abstract

No abstract