Welcome to QUENDS¶

QUENDS : Quantifying Uncertainty in ENsemble Data Streams, is a Python-based framework for analyzing time-series outputs from simulations and experiments. It helps transform raw, noisy data streams into reliable statistical summaries by identifying and trimming transient behavior, detecting steady-state regions, and estimating uncertainty in both single-run and ensemble simulation outputs.

QUENDS is designed for workflows where simulation outputs evolve over time and where trustworthy post-processing is needed before drawing conclusions. The framework supports single-trace analysis, ensemble-based analysis, steady-state detection, transient trimming, statistical estimation, and uncertainty quantification. By combining these capabilities into a unified workflow, QUENDS provides researchers with a reproducible way to move from raw simulation data to meaningful, uncertainty-aware results.

Quickstart¶

pip install quends

import quends as qnds

# load one signal (+ its time column), drop the warm-up, quantify
ds = qnds.from_csv("output.csv", "Q_D/Q_GBD")
trimmed = ds.trim(method="threshold", window_size=100, threshold=0.1)
print(trimmed.compute_statistics())   # mean + uncertainty (effective sample size)

Authors¶

QUENDS is developed and maintained by Abeyah Calpatura, Bert Debusschere, Evans Etrue Howard, and Pieterjan M. Robbe, together with contributions from the wider community. See the full author list.