View and compare parameter estimates

A list of parameter estimates is available in the Estimates tab in the right section of the Pirana win­dow.

ParameterEstSingleRun.png 

Through the use of brackets and color additional information is presented in the Estimates tab.

(RSE for parameters is shown in round brackets)

[Shrinkage for the random effects is shown in square brackets]

Parameters where the final gradients were zero have red text

Parameters were fixed have a gray background

Parameters where boundaries were encountered have a blue background

Parameters where the mean of eta-distribution was significantly different from zero include the etabar with a red background

Using the Estimates tab toolbar, you can easily access the following tools and functionality:

icon_paramestimates.png: Display the Parameter Estimates dialog, which gives a more detailed list of estimates.

icon_R_1.png: Export the estimates using R.

icon_genreport.png: Generate report.

icon_shownmoutput.png: View the NM output file.

icon_datainspector_2.png: Open dataset/output files in DataInspector

icon_viewmatrix.png: View matrix. Select type of matrix from the pop-up menu (see “Matrices” for more information).

The more detailed Parameter Estimates dialog can be accessed using the any of the following meth­ods:

Click icon_paramestimates_1.png in the Pirana toolbar.

Click icon_paramestimates_2.png in the Estimates tab toolbar.

Select Results > Show/compare parameter estimates.

Select View > Show parameter estimates.

Right-click a selected model in the model overview list and choose Model > Parameter estimates from the menu.

Right-click a selected model in the model overview list and choose Batch operations > Show parameter estimates from the menu.

ParameterEstSingleRunDetails.png 

If multiple models/runs are selected when the Parameter Estimates dialog is opened, the parameter estimates for all selected runs are presented, side-by-side, facilitating comparison.

ParameterEstimatesMultipleRuns.png 

The pull-down menu offers different choices for transforming the variances. Options include: Untransformed, (sqrt(OM^2), and (sqrt(exp(OM^2)-1).

ParameterEstSingleRunBottom.png 

Check the exp(THETA) box to back transform thetas that have been log-transformed in model code.

Note:Pirana will automatically back-transform when the parameter description includes the text “[log]”, e.g.:
    #THETA
    (0.,1.6,4.) ;KA [log]
 

From the dialog, convert results into CSV format, LaTeX, or HTML for reports or further analysis using the buttons along the bottom.

ParameterEstButtons.png 


Last modified date:12/17/20
Certara USA, Inc.
Legal Notice | Contact Certara
© 2020 Certara USA, Inc. All rights reserved.