Change: [Linkgraph] Improve distance scaling algorithm in demand scaler #12542
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Motivation / Problem
The current scaling algorithm produces poor quality results at values other than 0% and 100%.
At large values (e.g. > 200%) scaling is erratic does not produce results consistent with a large setting value.
This broadly is due to:
++chance > this->accuracy * num_demands * num_supplies
, etc) which leads to poor results.Description
Change the base distance value to IntSqrt(DistanceMaxPlusManhattan(map size)), which gives much more reasonable values and outputs, whilst still being responsive to variable map size.
Decrease the magnitude of the "Increase effect of mod_dist > 100" adjustment.
Change the scaling distance adjustment, such that it gives reasonable results with the now reduced range of mod_dist, and do not allow negative distances.
Scale divisor by a constant to reduce rounding losses.
The outputs are broadly the same as the existing algorithm at 0% and near 100%.
Limitations
N/A
Checklist for review
Some things are not automated, and forgotten often. This list is a reminder for the reviewers.