Allowed q > 1 and added unit tests to check magnification

[rapoliveira]

Removing two conditions of q <= 1 from modelparameters.py, one unit test was changed and two unit tests were added to ensure that the magnification is reproduced by the transformations q -> 1/q and alpha -> alpha +/- 180.

Parameters from example01 were used to test different mass ratios, except for the separation s, which was changed to reproduce a similar caustic crossing for binary lens.

See issue #84 for more details. Comments are welcome!

[rpoleski]

Few changes are needed:

• if new_q < 0.: ----> <= 0.
• correct error messages, so that they don't say q should be below 1.
• New tests look quite long (many epochs there). Please try to make them faster, i.e., spend a few minutes finding which ~5 epochs probe potentially hard cases (e.g., close to caustics) so that running tests will be faster and we will save developers time in the long run :)

[rpoleski]

The last added test should be corrected. Qualitatively the difference between q of 0.99 and 1.01 should be similar to difference between 0.97 and 0.99. You may choose a trajectory that is passing close to the caustic but is not crossing it.

[rpoleski]

Please change version to 2.24.0. I'll merge after that.