The indicator type. More...
Data Fields | |
| char * | name |
| Name of the indicator used for identification and the output. More... | |
| FILE * | dat_file |
| File for logging indicator values at predefined values. More... | |
| FILE * | tdat_file |
| File for logging indicator values at predefined evaluations. More... | |
| FILE * | info_file |
| File for logging summary information on algorithm performance. More... | |
| int | target_hit |
| Whether the target was hit in the latest evaluation. More... | |
| coco_observer_targets_t * | targets |
| Triggers based on target values. More... | |
| int | evaluation_logged |
| Whether the whether the latest evaluation was logged. More... | |
| coco_observer_evaluations_t * | evaluations |
| Triggers based on numbers of evaluations. More... | |
| double | best_value |
| The best known indicator value for this problem. More... | |
| double | current_value |
| The current indicator value. More... | |
| double | additional_penalty |
| Additional penalty for solutions outside the ROI. More... | |
| double | overall_value |
| The overall value of the indicator tested for target hits. More... | |
| double | previous_value |
| The previous overall value of the indicator. More... | |
Detailed Description
The indicator type.
The hypervolume indicator ("hyp")
The hypervolume indicator measures the volume of the portion of the ROI in the objective space that is dominated by the current Pareto front approximation. Instead of logging the hypervolume indicator value, this implementation logs the difference between the best know hypervolume indicator (a value stored in best_value) and the hypervolume indicator of the current Pareto front approximation (current_value). The current_value equals 0 if no solution is located in the ROI. In order to be able to register the performance of an optimizer even before the ROI is reached, an additional value is computed when no solutions are located inside the ROI. This value is stored in additional_penalty and equals the normalized distance to the ROI of the solution closest to the ROI (additional_penalty is set to 0 as soon as a solution reaches the ROI). The final value to be logged (overall_value) is therefore computed in the following way:
overall_value = best_value - current_value + additional_penalty
- Note
- Other indicators are yet to be implemented.
Field Documentation
| double additional_penalty |
Additional penalty for solutions outside the ROI.
| double best_value |
The best known indicator value for this problem.
| double current_value |
The current indicator value.
| FILE* dat_file |
File for logging indicator values at predefined values.
| int evaluation_logged |
Whether the whether the latest evaluation was logged.
| coco_observer_evaluations_t* evaluations |
Triggers based on numbers of evaluations.
| FILE* info_file |
File for logging summary information on algorithm performance.
| char* name |
Name of the indicator used for identification and the output.
| double overall_value |
The overall value of the indicator tested for target hits.
| double previous_value |
The previous overall value of the indicator.
| int target_hit |
Whether the target was hit in the latest evaluation.
| coco_observer_targets_t* targets |
Triggers based on target values.
| FILE* tdat_file |
File for logging indicator values at predefined evaluations.
The documentation for this struct was generated from the following file:
- C:/Users/dimo/Desktop/coco/code-experiments/src/logger_biobj.c
