Missing a minus sign in line 30 of unit3-param.cc? about maplab HOT 4 CLOSED

I'm not very familiar with maplab, but it looks like there was a normalization to positive w previously, then it was later removed (along with the commented out asserts).

To be honest I'm not sure why you would want to normalize the quaternion positive w for Jacobian computation in the first place (if someone thinks otherwise, please enlighten me). Without looking too much at the context, I think it should be fine as is, but the code could be cleaned up to remove quat_x_copy altogether, I guess.

from maplab.

@BITVoyager thanks for the report, this does indeed look fishy, we will look into it.

from maplab.

Looks like a bug, it should indeed have a minus. Thanks for reporting.

from maplab.

Fixed in the july release candidate.

from maplab.