Skip to main content

Features

GBSI-US Features are the derived inputs that underpin both Signals and Events.

They provide clean, research-ready representations of key components of the US gas balance.

What GBSI-US Features include​

Features may be derived from:

  • Supply and demand components
  • Storage dynamics
  • Flow and infrastructure behaviour
  • Weather-driven influences on consumption

All features are constructed using stable definitions and consistent transformations.

Since model 4.0 (3 October 2026), power_burn_bcfpd is the US national EIA series (gas consumed for power, Bcf/d), and the demand component uses its five-year same-month mean. Window-based columns such as balance_4w_momentum and balance_acceleration are computed over full history on every run.

Typical use cases​

  • Custom signal development
  • Quantitative modelling and screening
  • Historical research on balance conditions
  • Cross-period and cross-regime comparison

Accessing GBSI-US Features​

from snowtrail import Snowtrail
client = Snowtrail(api_key="your-api-key")

# Fetch GBSI-US features (e.g. balance momentum)
df = client.gbsi_us.balance_momentum(date_from="2024-01-01")

# All methods return pandas DataFrames
print(df.head())

GBSI-US Features are designed to integrate directly into quantitative research workflows.