CDC COVID-19 Forecast Hub: Weekly Probabilistic Forecasts of COVID-19 Hospital Admissions and Emergency Department Visits by U.S. State (Hubverse)
- Provider
- U.S. Centers for Disease Control and Prevention (CDC) and participating modeling teams
- License
- Public Domain (US Government work) (source)
- Format
- CSV (in sampled files)
- Stored at
- Amazon S3, s3://covid19-forecast-hub (us-east-1)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Health & Biomedical
- Source
- https://docs.hubverse.io/
About this dataset
The COVID-19 Forecast Hub, run by the U.S. Centers for Disease Control and Prevention (CDC), collects weekly probabilistic nowcasts and forecasts of COVID-19 activity from participating modeling teams, together with the observed target data used to score them. Teams forecast, for the United States, all 50 states, Washington DC and Puerto Rico, the weekly number of laboratory-confirmed COVID-19 hospital admissions reported to the National Healthcare Safety Network (NHSN) Hospital Respiratory Data, and the weekly percent of emergency department (ED) visits due to COVID-19 from the National Syndromic Surveillance Program (NSSP), for horizons from the previous epiweek to three weeks ahead. Forecasts are quantiles and, optionally, sampled trajectories, in the hubverse model-output schema. This public bucket is the hub's real-time mirror: model-output/ (one folder per team model, for example CovidHub-baseline), model-metadata/ (each model's description and data license), target-data/ (the observed time series), hub-config/ and auxiliary-data/; the raw/ folder is internal and should not be used. Use it for infectious disease forecasting research, ensemble methods and forecast evaluation (weighted interval score, coverage). Read with the hubverse hub-data (Python) or hubData (R) packages, pyarrow or pandas. License: the repository is a U.S. Government work in the public domain; each model's output carries the license named in model-metadata (CC BY 4.0, CC BY-SA 4.0 or CC0). Contact covidhub@cdc.gov about attribution of source forecasts.
How to access this dataset in Python
import boto3
from botocore import UNSIGNED
from botocore.client import Config
s3 = boto3.client('s3', region_name='us-east-1', config=Config(signature_version=UNSIGNED))
s3.list_objects_v2(Bucket='covid19-forecast-hub', Prefix='', Delimiter='/')Lists the files with anonymous access. Runs on any machine, with no cloud account.
Data structure
File types in sampled files: CSV
Files
Example file: auxiliary-data/weekly-model-submissions/2024-11-23-models-submitted-to-hub.csv
Analyze this dataset without downloading it
On Scigantic, CDC COVID-19 Forecast Hub: Weekly Probabilistic Forecasts of COVID-19 Hospital Admissions and Emergency Department Visits by U.S. State (Hubverse) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Health & Biomedical datasets
- UnumLocalia: same-section H&E, Xenium 5,001-gene spatial transcriptomics and COMET protein imaging of liver and tonsil
- Malaria Atlas Project: Plasmodium falciparum Malaria Death Count, 5 km annual maps
- Clinical Ultrasound Image Repository
- S-SYNTH: synthetic dermoscopy-style skin lesion images with segmentation masks (FDA)
- IDC GBM-DSC-MRI-DRO: Phantom MRI (3 patients)
- IDC MIDRC-RICORD-1B: COVID-19 (non-cancer) CT (117 patients)
- IDC Prostate-MRI-US-Biopsy: Prostate Cancer MRI, M3D, ultrasound (1,151 patients)
- 3D CyCIF Imaging of Tertiary Lymphoid Structures in Human Pulmonary Tuberculosis (Ogbonna et al. 2026)