A quick-and-dirty check for a one-dimensional active subspace

TL;DR

This paper introduces a rapid method for detecting a dominant one-dimensional active subspace in multivariate models by using a linear approximation, aiding in understanding input-output variability.

Contribution

It presents a quick, visualization-based check for identifying a primary active subspace without relying on statistical interpretation.

Findings

Enables fast detection of one-dimensional active subspaces

Provides a visualization tool related to regression graphics

Avoids complex statistical analysis for initial screening

Abstract

Most engineering models contain several parameters, and the map from input parameters to model output can be viewed as a multivariate function. An active subspace is a low-dimensional subspace of the space of inputs that explains the majority of variability in the function. Here we describe a quick check for a dominant one-dimensional active subspace based on computing a linear approximation of the function. The visualization tool presented here is closely related to regression graphics, though we avoid the statistical interpretation of the model. This document will be part of a larger review paper on active subspace methods.