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
- :
- Cluster Analysis

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
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Cluster Analysis

Mar 13, 2009 1:18 PM
(1825 views)

Is there a limit to the number of data points that can be analyzed using Analyze/Multivariate/Cluster? I have a set of 72 rows x 72 columns, and when I run a hierarchical clustering analysis, only about 25 of the rows are included.

3 REPLIES 3

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

Re: Cluster Analysis

I have used cluster with a lot more than 25 rows before, and the Statistics and Graphics Guide says that hierarchical clustering should work up to around 1,000 rows. I assume you haven't excluded the points that don't show up in your cluster and that they aren't missing data in the column(s) you are trying to cluster around.

Are there any other platforms that don't seem to be working right?

Are there any other platforms that don't seem to be working right?

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

Re: Cluster Analysis

There are missing data (which are just blank). Could this be the problem? I haven't excluded any of the rows or columns.

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

Re: Cluster Analysis

Do the blanks correspond with the rows that don't show up in your cluster analysis? If they do and the missing data is in columns that you are using for the clusters, that makes sense. If they match up but the missing data shouldn't be needed for the cluster analysis or if they don't match up at all, you may want to call tech support or post a bug on the knowledge base.