Subscribe Bookmark RSS Feed

How do I use [Validation Column] in [Validation Method] of Gen'zd (logistic) regression?

Oleg

Contributor

Joined:

Sep 13, 2017

My data has a column Train (values 1,0) to split the observations into training (=1) and validation (=0) for a Lasso Logistic Regression model. However, the "Validation column" option in the dropdown of "Validation Method" in the Run form of Generalized Regression is disabled, while Generalized Regression should have support of Validation Column (see https://www.jmp.com/en_us/software/predictive-analytics-software/key-features-of-jmp-pro.html). See printscreen.

 

How can I use this Validation Column (interactively at first) and where can I readh more about its use. Oddly, this window does not have help describing its components; and I failed to find anything online. Thanks in advance!

 

Screen Shot 09-27-17 at 06.46 PM.PNG

2 ACCEPTED SOLUTIONS

Accepted Solutions
txnelson

Super User

Joined:

Jun 22, 2012

Solution

Have you read the documentation on Generalized Regression in the "Fitting Linear Models" book?

     Help==>Books==>Fitting Linear Models

Jim
Oleg

Contributor

Joined:

Sep 13, 2017

Solution

Ok. In case others are stumbling over this, I figured out that validation column needs to be specified on Model Specification form in Validaiton field.

 

I would still love to know where how to get a descriptive help to buttons on any of the JMP window. It is common to have a [?] symbol at the top right corner of MS Windows forms, but JMP forms lack this feature. Plowing through books and web resources is not very helpful and time consuming (as is with this example, where the answer was not readily available). Any insight? 

 

The validation column must be indicated in advance on Model Specificaiton form. Place it to Validation input box.

Screen Shot 09-28-17 at 09.10 AM.PNG

7 REPLIES
txnelson

Super User

Joined:

Jun 22, 2012

Solution

Have you read the documentation on Generalized Regression in the "Fitting Linear Models" book?

     Help==>Books==>Fitting Linear Models

Jim
Oleg

Contributor

Joined:

Sep 13, 2017

Thank you Nelson. I love your enthusiasm about reading manuals. Yes, initially, as I looked through this particular book, I could not find the answer. It does elaborately describe how to create a validation column (and there is a beautiful online video too), but doesn't seem to go beyond this. Later I noted that p.37 describes elements of the Launch Window, including the Validation box. Thanks for your help.

Oleg

Contributor

Joined:

Sep 13, 2017

Solution

Ok. In case others are stumbling over this, I figured out that validation column needs to be specified on Model Specification form in Validaiton field.

 

I would still love to know where how to get a descriptive help to buttons on any of the JMP window. It is common to have a [?] symbol at the top right corner of MS Windows forms, but JMP forms lack this feature. Plowing through books and web resources is not very helpful and time consuming (as is with this example, where the answer was not readily available). Any insight? 

 

The validation column must be indicated in advance on Model Specificaiton form. Place it to Validation input box.

Screen Shot 09-28-17 at 09.10 AM.PNG

Jeff_Perkinson

Community Manager

Joined:

Jun 23, 2011


Oleg wrote:

 

I would still love to know where how to get a descriptive help to buttons on any of the JMP window. It is common to have a [?] symbol at the top right corner of MS Windows forms, but JMP forms lack this feature. Plowing through books and web resources is not very helpful and time consuming (as is with this example, where the answer was not readily available). Any insight? 

 

 

Screen Shot 09-28-17 at 09.10 AM.PNG


JMP has a number of ways of accessing Help. Most platform launch dialogs have a Help button. You can see it in the Fit Model dialog that you've shown above. It's right next to your cursor.

 

In report windows you can use the question mark tool available in the toolbar at the top of the window. (Can't see the the toolbar or menu? See the image below.)

 

You can also switch to the question mark tool by clicking the question mark on your keyboard. Click on anything in the report window with the question mark tool and you'll get help on it.

 

medium.gif

 

-Jeff
txnelson

Super User

Joined:

Jun 22, 2012

Please note.......taken directly from

     Help==>Books==>Fitting Linear Models

in the section entitled

     Generalized Regression Models

an example that details each item in the input dialog box

Example of Generalized Regression
The data in the Diabetes.jmp sample data table consist of measurements on 442 diabetics. The
response of interest is Y, disease progression measured one year after a baseline measure was
taken. Ten variables thought to be related to disease progression are also measured at
baseline. This example shows how to develop a predictive model using generalized regression
techniques.
1. Select Help > Sample Data Library and open Diabetes.jmp.
2. Select Analyze > Fit Model.
3. Select Y from the Select Columns list and click Y.
4. Select Age through Glucose and click Macros > Factorial to degree.
This adds all terms up to degree 2 (the default in the Degree box) to the model.
5. Select Validation from the Select Columns list and click Validation.
6. From the Personality list, select Generalized Regression.
7. Click Run.
The Generalized Regression report that appears contains a Model Launch control panel
and a Standard Least Squares with Validation Column report.
In the Model Launch control panel, note the following:
– The default estimation method is the adaptive Lasso.
– The Validation Method is set to Validation Column because you specified a validation
column in the Fit Model window.
8. Click Go.
An Adaptive Lasso with Validation Column report appears. The Solution Path report
(Figure 6.2) shows plots of the parameter estimates and scaled negative log-likelihood.
The shrinkage increases as the Magnitude of Scaled Parameter Estimates decreases. The
estimates at the far right of the plot are the maximum likelihood estimates. A vertical red
line indicates those parameter values selected by the validation criterion, in this case, the
holdback sample defined by the column Validation.

I believe that item 5 is what you were looking for

 

Jim
Oleg

Contributor

Joined:

Sep 13, 2017

Thanks Jim. This was helpful indeed!

Oleg

Contributor

Joined:

Sep 13, 2017

Thank you Jeff. I missed the Help button :) It is helpful indeed.