Not sure if I missed a cleaning step in my community setup, but today, while using cheddar v0.1.639, I encountered unexpected behavior in TrophicSpecies() / LumpTrophicSpecies() where nodes that share identical sets of prey and predators are not lumped into the same trophic species.
The issue appears to arise because TrophicSpecies() compares prey and predator vectors using order-sensitive equality rather than set equality. As a result, two nodes with identical prey/predator memberships but different internal ordering of those vectors are treated as distinct trophic species.
I found this issue because I was trying to write a sensitivity analysis of a community and discovered I would get very large differences in trophic species.
Again, I am not sure if there is a step earlier that forces the ordering.
Thanks.
Not sure if I missed a cleaning step in my community setup, but today, while using cheddar v0.1.639, I encountered unexpected behavior in TrophicSpecies() / LumpTrophicSpecies() where nodes that share identical sets of prey and predators are not lumped into the same trophic species.
The issue appears to arise because TrophicSpecies() compares prey and predator vectors using order-sensitive equality rather than set equality. As a result, two nodes with identical prey/predator memberships but different internal ordering of those vectors are treated as distinct trophic species.
I found this issue because I was trying to write a sensitivity analysis of a community and discovered I would get very large differences in trophic species.
Again, I am not sure if there is a step earlier that forces the ordering.
Thanks.