CFA STF Data¶

Data access helpers for the Center for Forecasting and Outbreak Analytics Short-Term Forecasting Team.

The cfa.stf.data package provides a consistent Polars interface to public forecasting inputs in the CFA data catalog:

• NHSN weekly hospital respiratory admissions
• NSSP daily emergency department visits
• generation interval, delay, and right-truncation probability mass functions
• NSSP tail-exclusion rules for incomplete or anomalous recent observations

Get started API reference

Design¶

The helpers expose explicit temporal controls. Observation dates select rows; as_of selects the catalog vintage available at a given date. Data-frame functions return a polars.LazyFrame by default so callers can extend and optimize the query before collecting it.

import datetime as dt

from cfa.stf.data import get_nssp

nssp = get_nssp(
    disease=["flu", "rsv"],
    state_abb="CA",
    as_of=dt.date(2026, 1, 15),
    start_date=dt.date(2025, 10, 1),
)

data = nssp.collect()

Package areas¶