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
- :
- Re: supress outliers

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

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

Mar 14, 2019 3:28 PM
(138 views)

I am trying to supress outliers in a continuous variable. In SPSS I used to just declaire a range and recode them as missing values. I know this should be obvious, but I need some help

thanks

1 ACCEPTED SOLUTION

Accepted Solutions

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

Still yet another technique from the JMP analysis platform report itself is to invoke a Local Data Filter. The nice aspect of this option is you can make real time changes right within the report itself to examine visually the effects of outlier manipulation.

7 REPLIES 7

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

Re: supress outliers

@emeritus ,

You can use the hide and select approach. Here try this:

dt = Open( "$SAMPLE_DATA/Big Class.jmp" ); // Open a sample data table

// Defining the range which you feel are outliers
dt << Select Where (:height < 62);

// Hiding and Excluding them from your analysis (they will remain in the data table, but are not included in your

//analysis
dt << Hide and Exclude;

// Clear your selection
dt << Clear Select;

// Perform a simple bivariate analysis - notice that the values for height < 62 no longer show up in your plot
dt << Bivariate( Y( :weight ), X( :height ), Fit Line );

Best

Uday

Uday

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

Re: supress outliers

Mar 14, 2019 4:29 PM
(128 views)
| Posted in reply to message from uday_guntupalli 03/14/2019 06:44 PM

Thanks, this looks pretty straight forward. The tutorial book shows examples of hide and select using the interface but I could not get it to go. The lasso tools looked fun but I couldn't get it to work. I think the bottom line here might be that I simply have to learn JMP scripting language.

Highlighted
##

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

Re: supress outliers

I would use the Column Property, "Range Check". It allows you to remove outliers on a column by column basis.

Jim

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

Re: supress outliers

I actually looked at that but got scared off because I did not totally understand it. I will give it a shot.

Thanks

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

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

Re: supress outliers

I am not quite there yet but I am aware of this journal function. Thanks for the tip, this sounds like a really powerful tool.

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

Re: supress outliers

JMP Journaling doesn't really have any relationship to JMP local data filters. Journaling is a JMP specific capability for creating and sharing what I like to call a 'lab notebook' for you JMP related work. Journals can come in very handy for sharing your work, keeping a running record of your JMP discoveries and findings. Journals can contain all sorts of information, JMP data, other files, pictures, links to external locations like files and web sites, etc. Journals are very utilitarian in their own rite.