Card: measurements in the pair rule, weekday pattern

README.md

@@ -62,13 +62,13 @@ Built from the daily snapshots of [cfahlgren1/hub-stats](https://huggingface.co/
 - `dl30` is the Hub's rolling 30-day download count. `dl_all` (all-time downloads) only exists from 2025-02-27, so exact daily downloads, computed as the difference of `dl_all` between snapshots, start on that date.
 - Some days are missing in the source (Aug 2024, Jun 2025, Apr 2026, May–Jun 2026). Totals are unaffected; daily values across a gap are averages.
 - A model or dataset is tracked once it has 10+ downloads in 30 days, 50+ all-time downloads, or at least one like. A Space is tracked once it has a like; the Hub doesn't publish visits for Spaces.
-- On some days the Hub's download counters stand still and catch up a day or two later, mostly on Wednesdays. Dataset counters freeze all at once; model counters often freeze only in part (newer repos stop, older ones keep counting), so the model total may only dip to half. A day is treated as a stall when the total is under 0.3 of its 15-day median, when at least 25% of the models with 60K+ monthly downloads didn't move at all, or when a dip under 0.7 is made up by the next or previous day. `hub_series` spreads each episode evenly over its days (sums unchanged), and the snapshots inside it are listed in `skip_days` (`meta.json` for models, `repos_meta.json` for datasets) so per-repo series can do the same. `frozen` in the same files holds the recent daily share of frozen counters.
+- On some days the Hub's download counters stand still and catch up a day or two later, mostly on Wednesdays. Dataset counters freeze all at once; model counters often freeze only in part (newer repos stop, older ones keep counting), so the model total may only dip to half. A day is treated as a stall when the total is under 0.3 of its 15-day median, when at least 25% of the models with 60K+ monthly downloads didn't move at all, or when a dip under 0.7 is made up by the next or previous day (a value spread over days without a snapshot counts as one measurement there, its days weighed in the sum). `hub_series` spreads each episode evenly over its days (sums unchanged), and the snapshots inside it are listed in `skip_days` (`meta.json` for models, `repos_meta.json` for datasets) so per-repo series can do the same. `frozen` in the same files holds the recent daily share of frozen counters.
 - Weekly figures (`dl_7d`, `likes_7d`, growth) are rates over the days the reference snapshots really span, so a skipped or missing snapshot never makes a week longer than 7 days. A repo with no reference snapshot shows no weekly figure, unless it is new that week. The growth rankings leave out repos whose week came mostly from a single day (over 60%, or 80% for repos created that month once they have three snapshots); `peak_share` holds that share.
 - When a repo's all-time counter hasn't moved in 31 days, `dl30` is set to 0 (the Hub sometimes keeps showing a stale 30-day count).
 - Twice (May 2025, June 2026) the all-time counters went down for a large share of models and came back days later. Those snapshots are in `skip_days` too, and downloads across them are measured from the last good snapshot to the first one after the counters recovered, so the recovery isn't counted as new downloads. A drop that doesn't recover within 8 snapshots is treated as a lasting correction by the Hub, not set aside.
 - A few snapshots hold only part of the repos (for models, 2025-11-24 with 289k of 1.13M). A snapshot with fewer than half the rows of the one before is left out: nothing is measured from or to it, the gap around it is spread like any missing day, and it is listed in `partial` and in `skip_days`.
 - Days without a snapshot share the per-day average of the next one. A stall window that reaches into such a stretch takes all of it, up to the snapshot that closes it, so a catch-up booked across missing days is spread together with the stall it belongs to.
-- After all this, a few days are still well under their local median and the days around them don't make up for it. `low_days` lists them: under 0.7 of the 15-day median (adjusted for the day of the week, since weekends run a little lower), in runs that the 3 days on each side don't make up by at least half. Days without a snapshot share one measurement with the snapshot that closes them, so they are judged together (2026-01-31, with no snapshot, and 2026-02-01 come out at 0.74 as one measurement: not low). They keep their measured values; the Hub chart greys them out instead of smoothing over them. The last two days are judged once their neighbours are known.
+- After all this, a few days are still well under their local median and the days around them don't make up for it. `low_days` lists them: under 0.7 of the 15-day median (adjusted for the day of the week, estimated from days with their own snapshot, since weekends run a little lower), in runs that the 3 days on each side don't make up by at least half. Days without a snapshot share one measurement with the snapshot that closes them, so they are judged together (2026-01-31, with no snapshot, and 2026-02-01 come out at 0.74 as one measurement: not low). They keep their measured values; the Hub chart greys them out instead of smoothing over them. The last two days are judged once their neighbours are known.
 - A few snapshots came well under a day after the one before, when hub-stats moved its collection time (2025-03-04 came 10.3 hours after 2025-03-03). The counters' daily update then lands in the next snapshot, so such a day is short, not low: `short_days` lists the snapshots taken under 18 hours after the previous one, they never count as low days, and `snapshot_at` keeps the time of the latest snapshot.
 - `uses.parquet` comes from the cards as they are today, dated by when each Space or model was created, not by when it started using what it lists.
 - Download counts follow the Hub's [counting rules](https://huggingface.co/docs/hub/models-download-stats); the Hub occasionally books delayed downloads on a single day.