Potential Bug in geodeticConversion

[Michael Mercurio]

I may have identified a potential bug in geodeticConversion::PCPF2LLA. The latitude is computed as:

llaPosition[0] = atan2( sqrt(pow(pcpfPosition[0],2.) + pow(pcpfPosition[1],2.)), pcpfPosition.norm());

I believe this should be:

llaPosition[0] = atan2(pcpfPosition[2], sqrt(pow(pcpfPosition[0], 2.0) + pow(pcpfPosition[1], 2.0)));

Is this function depreciated, or is a fix required?

Thank you!

[benjamin....@u-space.fr]

Hello, 

It appears that the unit tests of the groundLocation module (the only BSk module to ever call PCPF2LLA() at Update()) do not actually call this function, and as such this bug may have gone unnoticed until now. 

[benjamin....@u-space.fr]

Sorry, i pressed send to quickly! 

In summary, a fix is required, as well as an augmented unit test.

[Hanspeter Schaub]

Howdy everyone, thanks for bringing this up.  Beginning of the semester here and I've not had a chance to confirm and create a fix.  It is on my to-do list.

[Hanspeter Schaub]

Howdy Michael and Ben, had a chance to look at this and yes, this `PCPF2LLA()` computed the latitude incorrectly.  I think the author was thinking of using the `acos()` function looking at the attempted math, but it was not correct.  I'm pushing a fix to develop and added a not to the "known issues" page.  This only impacts users who used the helper function `specifyLocationPCPF()` method inside the `groundLocation` module.