This release includes small bug fixes, re-organized LUT naming for more stable reuse of radiative transfer runs, and enhanced apply_oe capabilities.
This release is backwards compatible, though may cause LUT recalculations in some circumstances.
2.0.1
This is a minor upgrade to ensure that lookup table files are always given consistent names across multiple runs. This enables re-using and sharing of precalculated lookup tables.
2.0.0
-Enables coupled VSWIR TIR inversions -Brings in substantial multicore speedups -Removes numba -General code improvements and stability -Configuration radiative transfer file format has been modified slightly, and are no longer strictly backwards compatible with v1.*
1.4.3
New faster, stable matrix inversions courtesy Jouni Susiluoto Removed vestigial VectorInterpolatorJIT code Fixed a recently-introduced array indexing issue in multidimensional lookup tables
1.4.2
* Closes PR 62 * Includes a small change to the VectorInterpolator class that dramatically improves its performance (~15x) * Closes PR 63 * Minor bug fixes
1.4.1
Surface reflectance partial derivatives: Incorporates a small but important fix to some partial derivative calculations in the radiative transfer codes.