Curve Fitting Toolbox™ provides an interactive app and command line functions for fitting curves and surfaces to your data. The toolbox enables you to explore relationships between data interactively, generate predictive models, and conveniently use or share your curve fits for postprocessing analysis.
The Curve Fitter app lets you load data from the MATLAB® workspace and choose from a variety of prebuilt models for regression, interpolation, and smoothing workflows. For complex parametric models, use the Custom Equation tab to tailor fit to your data. Assess and validate the quality of your fitted models with goodness-of-fit metrics such as R-squared and root mean squared error. Additionally, visual analysis of residual plots provides another layer of insight into model performance. To further enhance the quality of your fits, you can tweak advanced fitting options or interactively exclude outliers.
Once you are satisfied with the quality of the fit, you can easily generate MATLAB code from the app and repeat the analysis with a new data set or batch process a large number of data sets. You can also export the fitted model to the MATLAB workspace to facilitate postprocessing analysis including prediction and forecasting, calculating integrals and derivatives, and estimating confidence intervals.
If you have Simulink® models, you can easily incorporate your fitted curves or surfaces into those models by exporting your fits as Simulink Lookup Table blocks directly from the app.
