Fix azimuth computation

[ChristopherRabotin]

Summary

The computation of the azimuth was off by 180 degrees. The algorithm is from the same source: Vallado, section 4.4.3.

How did this bug come to be?

Vallado specifies that the azimuth needs to be computed by accounting for the quadrant check, and provides the sine and cosine of the azimuth angles as follows:

• \sin \beta = \rho_y / norm(\rho)
• \cos \beta = -\rho_x / norm(\rho)

Given the quadrant check, I rewrote this as:

• \tan \beta = \sin \beta / \cos \beta = \rho_y / (-\rho_x) = -\rho_y / \rho_x

This is mathematically correct. However, when using the atan2 function, the sign of each component matters. The code was using atan2 incorrectly because I had moved the negative sign from the denominator to the numerator.

Architectural Changes

No change

New Features

No change

Improvements

No change

Bug Fixes

• Azimuth computation was wrong by 180 degrees.

Testing and validation

There isn't yet any public validation tests for azimuth and elevation, but it's been checked against proprietary data, which uncovered the error in azimuth. Elevation and range matched the proprietary data with high precision.

Documentation

This PR does not primarily deal with documentation changes.