Solve Global Least Squares dialog box |
![]() ![]() ![]() |
Use this dialog box to solve static least squares solutions.
Options
Select this option to display the joint error of all outputs.
Select this option to display a single output error.
Select one of the available error types:
Joint Error Types
Mean Square
Indicates the display of the mean square error:
.
Mean Absolute
Indicates the display of the mean absolute error:
.
Mean Distance
Indicates the display of the mean Euclidean distance between the output vector and the target vector:
.
![]() | For N=1, the Mean Absolute and the Mean Distance errors are equal. |
Single Error Types
Square
Indicates the display of the square error:
.
Absolute
Indicates the display of the absolute error:
.
Residuals
Indicates the display of the residual error:
.
![]() | In the preceding equations, Ti corresponds to the ith target value and Oi to the ith output value. |
Select from this list the output number for which you want to display approximation errors.
![]() | This option is available when Single Errors is selected. |
Each output number corresponds to a dependent model variable in the same order of definition.
Refer to Data and System Models for additional information on this subject.
Statistics
Displays the mean of the computed errors.
Displays the coefficient of determination for each of the outputs, calculated as:
Where is the coefficient of determination for the ith output,
is the ith. output for the jth sample,
is ith target value at the jth sample, and
is the mean of the ith target value for all considered samples.
![]() | The coefficient of determination is only displayed for single error types. |
Displays the minimum of the computed errors.
![]() | When Residuals is specified for the Error Type option, this box displays the minimum error value, as opposed to the minimum error. |
Displays the maximum of the computed errors.
![]() | When Residuals is specified for the Error Type option, this box displays the maximum error value, as opposed to the maximum error. |
Plots approximation errors corresponding to the selected error type.
Refer to Mini-Plot box shortcut commands for additional information on the use of mini-plot boxes.
![]() | Use the Extract To command on the Mini-Plot box shortcut menu to extract the values corresponding to the data points of the visible chart series into a Predictive Systems Lab document worksheet. |
Specifies the color with which the series corresponding to the calculated error values is plotted. Click this box to change the current color.
Specifies the color with which the chart series corresponding to the calculated error running average is plotted. Click this box to change the current color.
The error running average at any time t is calculated as
.
Displays the Global Least Squares Options dialog box. Click this button to review or modify static area settings.
Reloads the least squares settings to the last values applied. Click this button to reload the least squares settings.
![]() | Reloading the settings to the last values applied erases the current least squares state. |
Solves the currently defined least squares problem. Click this button to solve the currently defined least squares problem.
Applies the current least squares solution to the solution method. Click this button to apply this solution to the method.
Prompts you to save any changes and closes the dialog box. Click this button when you have finished solving the least squares problem.
Try Predictive Systems Lab
Need Help?
Contact support