Spreadsheet Debugging

TL;DR

This paper introduces an efficient fault tracing technique for spreadsheets, leveraging interval-based testing and dataflow analysis to identify influential faulty cells, thereby improving debugging of spreadsheet programs used for numerical computations.

Contribution

It presents a novel fault tracing method based on interval testing and dataflow analysis, enhancing debugging capabilities for spreadsheet artifacts.

Findings

Effective identification of influential faulty cells

Improved debugging efficiency for spreadsheet errors

Utilizes dataflow and cell marking for fault localization

Abstract

Spreadsheet programs, artifacts developed by non-programmers, are used for a variety of important tasks and decisions. Yet a significant proportion of them have severe quality problems. To address this issue, our previous work presented an interval-based testing methodology for spreadsheets. Interval-based testing rests on the observation that spreadsheets are mainly used for numerical computations. It also incorporates ideas from symbolic testing and interval analysis. This paper addresses the issue of efficiently debugging spreadsheets. Based on the interval-based testing methodology, this paper presents a technique for tracing faults in spreadsheet programs. The fault tracing technique proposed uses the dataflow information and cell marks to identify the most influential faulty cell(s) for a given formula cell containing a propagated fault.