- Improve benchmark script stability - Fix mutual information metric in baselines - Update baselines to use transforms
1.0.6
Key Updates
- Add Adaptive-IPP benchmark and example notebook - Update non-point FoV transforms to be compatible with IPPTransform and also usable as standalone transforms - Change the default optimizer to Scipy
1.0.5
Key Updates
- Fix candidate points scaling issue in the data preprocessing module - Change default optimizer to Scipy - Set IPPTransform continuous sampling type to w/o FoV aggregation - Improve benchmark script stability
1.0.4
Key Updates
- Update dataset preprocessing utility to use PIL instead of GDAL to read tif files - Reduce package dependencies to pip installable packages
1.0.3
Key Updates
- Add online/adaptive IPP benchmark - Validate tensorflow 2.17.0 compatibility - Update genetic algorithm-based IPP method to support sgp-tools transforms
1.0.2
- Update TSP solver with args to define optional start and end locations for the solution path(s) - Update the gpflow optimizer with args to select the optimization method with a string. The method has to be one of the available methods from scipy minimize and tensorflow optimizers