Classification trees for the planet-harboring status among dwarf stars in our sample. The nodes are annotated with a categorical binary variable that assumes values 1 or 0 for stars with and without a giant planet, respectively. The leaves are marked with statistics of the partitioned group: fraction of d1 stars in the group (stars harboring a giant planets) and percentage of all dwarfs stars that were classified in that leaf. The top left tree is the full partition model for our sample. The other trees consider the case when two important elements are not known: V and Cu, one at a time or both.