Julian Pawlowski 64cf842719 fix(rating): improve gap tolerance to find dominant large blocks ... The gap tolerance algorithm now looks through small intermediate blocks to find the first LARGE block (> gap_tolerance) in each direction. This ensures small isolated rating intervals are merged into the correct dominant block, not just the nearest neighbor. Example: NORMAL(large) HIGH(1) NORMAL(1) HIGH(large) Before: HIGH at 05:45 merged into NORMAL (wrong - nearest neighbor) After: NORMAL at 06:00 merged into HIGH (correct - dominant block) Also collects all merge decisions BEFORE applying them, preventing order-dependent outcomes when multiple small blocks are adjacent. Impact: Rating transitions now appear at visually logical positions where prices actually change direction, not at arbitrary boundaries.		2025-12-22 13:28:25 +00:00
..
tibber_prices	fix(rating): improve gap tolerance to find dominant large blocks	2025-12-22 13:28:25 +00:00