Linear Regression
Origin's linear regression enables you to perform a linear fit
on your data (y=A+Bx), where A is the intercept and B is the slope.
You can automatically perform linear regression through Origin's
Analysis menu or set up a Linear fit using the Linear Fit tool.
Initial parameter estimates for each linear fit are calculated automatically
and the results of the fit are reported to the Results Log.
The results include the parameter values, standard errors, correlation
coefficient (R), standard deviation, number of points used in the
fit (N), and the p-value (probability that R=0).
|
|
The Settings tab:
Perform an Apparent Fit when your data appears
linear on another axis scale (e.g. log). This might be especially
useful when attempting to determine decay constants.
Extrapolate the Linear Fit by selecting Span X Axis
check box or by defining a percent value in the Range
Margin text box.
Generate Residual Data in your original worksheet.
Fit All Curves simultaneously in a graph window.
Obtain Simple or Advanced statistics.
- Simple Statistics includes intercept and
slope values, standard error values for each parameter,
R (the correlation coefficient), the standard deviation,
the P value for the t-test of the slope = 0, and more.
- Advanced Statistics includes Simple Statistics
as well as t-test values and the ANOVA table.
|
 |
The Operation tab:
Force the linear fit through the Origin (Through Zero).
Fix the Slope at a particular
value.
Use an Error Dataset as Weighting.
Include Confidence Bands and/or Prediction
Bands.
Find X given a Y value in the fit curve or
Find Y given an X value in the fit curve in
the Calculate group.
|
To learn more about curve fitting in general, read the "Primer"
called "Curve
Fitting made easy" that was written by one of our senior
developers for the Industrial Physicist web site.
|
