five

EPI-Eval/flu-metrocast-hub

收藏
Hugging Face2026-04-26 更新2026-05-03 收录
下载链接:
https://hf-mirror.com/datasets/EPI-Eval/flu-metrocast-hub
下载链接
链接失效反馈
官方服务:
资源简介:
--- pretty_name: Flu MetroCast Hub — sub-state flu hosp forecast target license: cc-by-4.0 size_categories: - 10K<n<100K tags: - cadence-weekly - geo-us - surveillance-respiratory - pathogen-influenza - tier-2 - availability-open schema_version: '0.1' source_id: flu-metrocast-hub source_url: https://reichlab.io/metrocast-dashboard/ manifest_section: §4.6 surveillance_category: respiratory pathogens: - influenza availability: open availability_notes: null access_type: github tier: 2 cadence: weekly geography_levels: - subnational-city geography_countries: - US gold_standard_for: [] vintaged_version_of: null succeeds: null derived_from: [] value_columns: - name: value unit: admissions/week value_type: incident description: Flu hospital admissions / ED visits for the metro area aggregation: sum notes: extra_columns: - column: target description: Hubverse target identifier (e.g. 'wk inc flu hosp metro'). - column: location_name description: MetroCast metro identifier (city / county). `location_id` is `US-METRO-<id>` synthetic. interpretation_caveats: - column: value caveat: 'Metro definitions vary by jurisdiction (NYC = 5 boroughs; LA = LA County; etc.). Don''t sum metros to recover state — they don''t tile cleanly. ' general: 'Truth-only ingest, model forecasts handled separately. ' extra: schema_proposal: '`subnational-city` location level needed a synthetic prefix (`US-METRO-`) because the schema''s regex for that level is permissive but un-anchored to a registry. Worth formalising: maintain a `metro_registry.yaml` with city codes and human names. ' computed: last_ingested: '2026-04-26T12:48:50Z' row_count: 132242 time_coverage: - start: '2024-08-03' end: '2026-04-18' geography_unit_count: 77 observed_cadence_days: 7 missing_gaps: [] data_hash: 9b5355d031f000d2 --- # Flu MetroCast Hub — sub-state flu hosp forecast target Truth-only ingest, model forecasts handled separately. **Source:** <https://reichlab.io/metrocast-dashboard/> ## Coverage - **Time:** 2024-08-03 → 2026-04-18 - **Cadence:** `weekly` (observed median spacing: 7 days) - **Geography levels:** `subnational-city` — 77 unique location IDs - **Countries:** US - **Pathogens:** `influenza` - **Surveillance category:** `respiratory` - **Rows:** 132,242 ## Columns | Column | Unit | value_type | Aggregation | Description | |--------|------|------------|-------------|-------------| | `value` | admissions/week | `incident` | `sum` | Flu hospital admissions / ED visits for the metro area | ### Additional data columns - **`target`** — Hubverse target identifier (e.g. 'wk inc flu hosp metro'). - **`location_name`** — MetroCast metro identifier (city / county). `location_id` is `US-METRO-<id>` synthetic. ## Interpretation caveats Things that may differ from how other sources define a similar measure. If you're combining this dataset with another, read these first. - **`value`** — Metro definitions vary by jurisdiction (NYC = 5 boroughs; LA = LA County; etc.). Don't sum metros to recover state — they don't tile cleanly. ## Access - **Availability:** `open` - **Access type:** `github` - **License:** cc-by-4.0 - **Tier:** 2 --- *Schema version `0.1` · Last ingested 2026-04-26T12:48:50Z · `source_id: flu-metrocast-hub` · Manifest section §4.6*
提供机构:
EPI-Eval
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作