4.2.2.9 Fitting With Multiple Independent Variables



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


Summary

The Function Organizer tool can be used to create user-defined functions with more than one independent or dependent variable. The NLFit dialog can then be used to fit with such functions. The preview window in the NLFit dialog is capable of plotting only one quantity versus another. However, even if the preview does not "make sense", the fitting process will correctly proceed, once proper data and parameter assignments have been made.

Note that if you wish to fit multiple independent variables with an equation of the type
y = A0 + A1 * x1 + A2 * x2 + ...
you can make use of the Multiple Regression tool instead of the nonlinear fitter dialog.

Minimum Origin Version Required: Origin 8.0 SR6

What you will learn

This tutorial will show you how to:

  • Create a user-defined fitting function with two independent variables and one dependent variable
  • Fit with that function in NLFit

Steps

  1. Start with a new workbook and import the file \Samples\Curve Fitting\Activity.dat.
  2. Select Tools: Fitting Function Organizer from menu (or press F9) to bring up the Fitting Function Organizer and define a new fitting function named MultiIndep in NewCategory (create the category if it does not exist) as follow:
    Function Name:

    MultiIndep

    Function Type:

    User-Defined

    Independent Variables:

    substr,inhib

    Dependent Variables:

    act

    Parameter Names:

    ki,km,vm

    Function Form:

    Origin C

    Function:

    double mix = inhib / ki;
    act = vm * substr / (km + (1 + mix) * substr);

    NOTE: Since we are using OriginC, case must match between defined names and their use in the function definition. e.g. Substr does NOT equal substr.

    Click Save and then OK to save the function and close the Organizer.

    For more details about User Defined Fitting Function please refer to User Defined Fitting Function using Origin C.

  3. Highlight ONLY column C and select Analysis: Fitting: Non-linear Curve Fit from menu to bring up the NLFit dialog. Select the function MultiIndep from NewCategory on the Settings: Function Selection page. Set the input datasets in the Data Selection page as follow:
    MultiFittingDataSelection1.png
  4. Select the Fitted Curves page and expand the Fitted Curves Plot branch. Then select Same as Input Data from the drop-down list next to the X Data Type branch.
    MultiFittingFittingCurves1.png
  5. Select Parameters Tab and set the initial values as follow:
    MultiFittingSettingPara1.png
  6. Click Fit button to generate the fitting reports. You can see these results from the report worksheet as below:
    MultiIndepParameter.png
    MultiIndepStat.png
    From the Statistics table we can know that the fitting is fairly successful.