What problem does it solve?
Choosing and refining a curve or surface fit often requires exploring many fit types, excluding outliers, and comparing goodness-of-fit metrics, which is tedious and error-prone when done blindly from the command line. This Skill guides an AI agent through the complete interactive fitting workflow in the MATLAB Curve Fitter app, from data selection to export.
Core Features & Use Cases
- Interactive fitting workflow: Select fitting and validation data, choose from regression, interpolation, smoothing, and custom equation fit types, adjust fit options, and evaluate results with goodness-of-fit metrics.
- Outlier handling and comparison: Exclude points interactively or via rules, duplicate fits to compare approaches side by side, and review residuals, prediction bounds, and confidence intervals.
- Export and sharing: Export fits to the MATLAB workspace, generate MATLAB code, save sessions as .sfit files, and export fits to Simulink lookup tables.
- Use Case: An engineer with sensor data in workspace variables asks the agent to open Curve Fitter, compare a polynomial and an exponential fit, exclude anomalous readings, and export the chosen fit as a Simulink lookup table.
Quick Start
Ask the agent to open the Curve Fitter app with your x and y workspace variables and fit a polynomial curve to the data.