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
- :
- Simple Regression Help

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

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

Simple Regression Help

Oct 10, 2018 1:16 PM
(725 views)

Hello!

I am having trouble when using fit model on a simple regression. Two variables do not work as seen below. They are simple numbers so I can't figure out why my parameters look the way they do. Any suggestions? Thanks so much.

3 REPLIES 3

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

Re: Simple Regression Help

Your two problem variables have the nominal modeling type. Therefore, each level of those variables (except one) will require a parameter estimate. This results in you not having enough data to estimate all of those parameters.

Those columns look to be numeric. I think you need to change the modeling types to be continuous and re-run your model. Then the results will likely be what you expect.

Dan Obermiller

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

Re: Simple Regression Help

Oct 10, 2018 2:05 PM
(710 views)
| Posted in reply to message from Dan_Obermiller 10/10/2018 04:34 PM

As follow up to what Dan says, you probably should just change those variables from nominal to continuous and rerun the regression.

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

Re: Simple Regression Help

Further agreeing with the reply from Dan, I would like to offer a suggestion before you switch modeling type to Continuous.

Do a 'Distribution' graph on the '% Discount' and 'Trough' columns, and see what values are present ... it is possible that something is present in the columns you don't want to lose but that makes the column appear non-numeric. Just changing to continuous will remove all non-numeric cell data.

Do a 'Distribution' graph on the '% Discount' and 'Trough' columns, and see what values are present ... it is possible that something is present in the columns you don't want to lose but that makes the column appear non-numeric. Just changing to continuous will remove all non-numeric cell data.