Provide tests for R/T by opaque sample
R/T calculations may become numerically instable if overall transmission approaches floating-point underflow. Considerable effort has been invested to get this right - yet it needs to be demonstrated and ensured by dedicated tests, for the scalar and the polarized case.
These tests shall then be used to validate the recent refactoring around commit 9ce0c2ba in git-main, where I removed an ad-hoc underflow adjustment.
Depends on #551, computation of transmission.
Which depends on #555 (closed), flatten simulation results.
Which depends on #71 (closed), refactor ICoordSystem. Which is related to a number of other issues.
Edited by Ammar Nejati