Skip to content

Added a missed patch inside test_contacts (as in !126)

In !126 (merged), it was added a patch that solves the numerical stability during NumDiff for a random vector. However it was missed to force the state values for next test.

Please note that an explanation of this numerical stability can be found in #139 (closed).


I detected this problem when I was running several times this unit-test. Now it's working well!

Merge request reports