Closed Shushman closed 8 years ago
These look like a good start.
However, they are very similar to the planner-agnostic tests we have in prpy
. Ideally, you should integrate your improvements to these tests into the prpy
tests, when possible. This will let us benefit from them on all of our planners that support these methods. Any tests that are planner-specific (e.g. test trajopt
-specific functionality) should stay standalone.
Sorry if this was misleading. We typically have two types of tests for planners:
prpy
wrapperUnlike most of our planners, we did not write the core C++ code. Therefore, we may only need to test (2).
@lgw903 added Trajopt tests to PrPy in https://github.com/personalrobotics/prpy/pull/213.
@Shushman Can you take a look at that pull request and add any tests that you think are missing? In particular, I definitely want to verify that we are testing the two bugs that you found.
A first attempt at tests for OR_TRAJOPT. A summary of the file:
PlanToConfiguration
,PlanToEndEffectorPose
,PlanToIK
,PlanToEndEffectorOffset
. Using the test values fromor_ompl
.PlanToTSR
as I need to figure out TSRs better.Issues to note: