I have a nicely tidy dataset of shrimp populations in several rainforest pools (in Puerto Rico) with several categorical columns: year, shrimp species and pool location. The primary numerical variable is “shrimp per trap” and it is measured weekly, so there is also a “week” variable. I’m often in the situation of wanting to graph the distribution of weekly measurements for a particular year and a particular species. I tend to drag the “year” to the left, select a year that way, set aside the unselected cases, then make a graph of all weekly measurements for that year, color it by shrimp species, select one species, then hide the unselected points. This works OK in terms of getting the graph I wanted, but I can no longer see which year I’ve selected because the species selection is showing up in the table, rather than the year selection. I hope the attached screen shot helps you visualize the situation.
Any suggestions? Would I be better off moving both year and species to the left? Make both selections in the table, rather than one in the table and one in the graph? This is related to a more general issue I have about keeping track of which points are being displayed/which are set aside, since CODAP doesn’t track or display the filters that have been applied.