Turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- Fitting a model which includes both a column parameter and a table variable

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Fitting a model which includes both a column parameter and a table variable

Dec 5, 2016 12:35 PM
(1323 views)

Would like to learn how to fit a model which includes not only a column parameter Aa, but a table variable Xx.

The model column formula references a column which depends on a table variable (and other data columns) in a complicated way.

I want to find combination of model column parameter Aa and table parameter Xx which best fits a data column.

Hope this isn't too pedestrian a question.

2 REPLIES 2

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Fitting a model which includes both a column parameter and a table variable

Dec 5, 2016 12:47 PM
(1315 views)
| Posted in reply to message from galactus3000 12/05/2016 03:35 PM

If the model equation is already in a column formula, then use the **Nonlinear** platform. The default loss function is least squares. First, though, replace the table variable in the formula with a formula parameter. In the formula editor, click **Table Columns** and select **Parameters**. Click **New Parameter**, give it a name **Xx**, and a **starting value**. Select the table variable in your formula and then select the new **parameter Xx** to replace it.

All of the model parameters must be entered like this one. The Nonlinear platform does not provide model parameters automatically like the Fit Least Squares platform, in which you supply the predictors as columns and JMP supplies the linear parameters.

Learn it once, use it forever!

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Fitting a model which includes both a column parameter and a table variable

Thank you for answer. Something I would do if I could.

The formula column involves another formula column. How do I get the Nonlinear model to see that other nested parameter?

The complication here is that I am performing a Col Cumulative Sum on a column that involves this second parameter Xx.

So model formula in Column 6 with parameter Aa references Column 2 which is a Col Cumulative Sum of Column 1 with formula involving parameter Xx.

Want to do a two-parameter fit here for parameters Aa and Xx.