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

Showing results for

- JMP User Community
- :
- Discussions
- :
- Profiler Set-Up: Output Variable (Y) Importance

- 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

Nov 7, 2017 8:47 AM
(486 views)

In multiple y optimization in a profiler, each y requires it's own desirability function/response goal (max, min, target). But to combine them into a single desirabililty function you have to assign an Importance value to each y. The default value = 1. How does this work? What is the scale?

Thanks, in Advance,

Steve

1 ACCEPTED SOLUTION

Accepted Solutions

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

Nov 7, 2017 8:56 AM
(963 views)

Solution

The importance does not NEED to be assigned, since the default is 1. However, the importance is just that: importance. So you can determine the scale. Suppose you have Y1 and Y2. If Y1 has an importance of 1 and Y2 has an importance of 2, then Y2 is twice as important as Y1. You would get exactly the same result if you specified importances of 0.5 and 1, 4 and 8, 100 and 200. The scales are relative.

Dan Obermiller

1 REPLY

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

Nov 7, 2017 8:56 AM
(964 views)

The importance does not NEED to be assigned, since the default is 1. However, the importance is just that: importance. So you can determine the scale. Suppose you have Y1 and Y2. If Y1 has an importance of 1 and Y2 has an importance of 2, then Y2 is twice as important as Y1. You would get exactly the same result if you specified importances of 0.5 and 1, 4 and 8, 100 and 200. The scales are relative.

Dan Obermiller