Problem
The optimizer derives N_g from --dataset (CSV series inventory) and uses one global placeholder arrival rate --rho (#563). Subpopulation-aware sketches are costed with a constant SUBPOPULATION_COUNT.
Change
Accept externally provided label-set facts per (metric, spatial filter, grouping labels): cardinality (distinct label-value combinations) and arrival_rate_per_sec (aggregate items/sec), as in sketch-bench#129's LabelSetInfo. Loaded from an input file; the optimizer does not estimate them.
- Replaces
--dataset / --rho on the optimizer path.
- Ingest cost uses the per-label-set arrival rate; group multipliers use the provided cardinality (including subpopulation-aware sketches).
- Missing facts for a needed label set → error.
Refs #693, #563, #525.
Part of #753.
🤖 Generated with Claude Code
Problem
The optimizer derives N_g from
--dataset(CSV series inventory) and uses one global placeholder arrival rate--rho(#563). Subpopulation-aware sketches are costed with a constantSUBPOPULATION_COUNT.Change
Accept externally provided label-set facts per (metric, spatial filter, grouping labels):
cardinality(distinct label-value combinations) andarrival_rate_per_sec(aggregate items/sec), as in sketch-bench#129'sLabelSetInfo. Loaded from an input file; the optimizer does not estimate them.--dataset/--rhoon the optimizer path.Refs #693, #563, #525.
Part of #753.
🤖 Generated with Claude Code