4.2.2.3 Global Fitting with Parameter Sharing


Video Image.png Video Text Image.png Website blog icon circle.png Blog Image 33x33px.png

Summary

The term "global fitting" generally refers to simultaneous curve fitting operations performed on multiple datasets. Because datasets remain distinct, they may or may not "share" parameter values during the fit process. When a parameter is shared, a single parameter value is calculated for all datasets; When a parameter is not shared, a separate parameter value is calculated for each dataset.

By contrast, "concatenated fitting" is performed by combining all datasets into a single dataset. Because curve fitting operations are performed on a single dataset, only a single set of parameter values is returned.

Since 2016 SR2, an new app Sequential Fit has been released to do sequential fitting on multiple datasets. With this app, the fit parameter values obtained from the current dataset will be used to initialize the parameters of the next dataset. This is very applicable for those datasets, the common parameters of which keep changing in turn.

What you will learn

This tutorial will show you how to:

  • Select multiple datasets for fitting.
  • Select different fitting modes.
  • Perform global fit with parameter sharing.

Steps

  1. Create a new workbook by clicking the Button New Workbook.png button.
  2. Click the Button Import Single ASCII.png to import the file Enzyme.dat in the <Origin EXE Folder>\Samples\Curve Fitting\ folder.
  3. Highlight column B and C and click the Button Scatter.png button to generate a scatter plot.
  4. Keep the graph window activated and select the Analysis: Fitting: Nonlinear Curve Fit menu item to open the NLFit dialog.
  5. Set the Recalculate mode to Auto.
  6. In the Function Selection page, choose the Hill function from the Growth/Sigmoidal category.
  7. Go to the Data Selection page, click the triangle button next to the Input Data selection box and choose Add All Plots in Active Layer, to add both plots as input data.
  8. Select Global Fit mode from the Multi-Data Fit Mode drop-down list:
    Tutorial Global Fitting 002.png
  9. Go to the Parameters tab. Check the Fixed check box for n and n_2 to fix their values to 1.
    Tutorial Global Fitting 003.png
  10. Click the Fit button to fit curves. You will get these results from the Parameters table of the report worksheet:
    Tutorial Global Fitting 005.png
  11. The fitted curves will be added to the original data plot.
    Tutorial Global Fitting 004 2.png
  12. In this case, the maximum velocity Vmax may be the same, so we could share this parameter value when fitting.
  13. Click the green lock icon in the report worksheet and select Change Parameters to bring back the NLFit dialog.
    Tutorial Global Fitting 010.png
  14. In the Parameters tab, check the Share check box for Vmax.
    Tutorial Global Fitting 008.png
  15. Click the Fit button again to update results. You can see the Vmax values for both curves are the same. The asterisk in the parameter name indicates that this parameter is shared:
    Tutorial Global Fitting 015.png
  16. The fitted curve will automatically update.
    Tutorial Global Fitting 016 2.png