MicroLensingDemo

On the left is a map of the inverse magnification. White contours are the critical curves. On the right is part of an image of a lensed Gaussian source. The convergence in stars is 1 for both.

This is a simple program to demonstrate how to do microlensing calculations with GLAMER.

There are more sophisticated things you can do with a Grid instead of a GridMap as used in this example. A Grid can be dynamically refined near the caustics or around the source images. But if you have enough memory and enough cores the GridMap works just fine.

In your simulation you should make sure:

1. You have set the N_THREADS compiler flag to as many cores as you can spare (cmake .. -DN_THREADS=20 in the build directory of GLAMER for example).
   This code is highly parallelized and will be much faster with more cores.

2. You have made a GridMap or Grid with a high enough initial resolution to resolve all the relavent images. This usually requires a resolution smaller than the Einstein ring radius of the stars.

3. You don't move the source far enough that the images are no longer on the Grid. You can always construct another grid centered on another point along the sources path and do it in stages. If there is a background shear it might be more efficient to have e rectangular Grid.

Good luck! Let me know about what you do with it.