cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Choose Language Hide Translation Bar
dainius
Level II

Labelling data points in PCA Biplot by a characteristic

Dear all,

 

is it possible to label the data points in a biplot shown in B by some definite characteristics in JMP? 

I mean, is there an option to label them by sample names from a different column or something, like I did it in A plot myself (in the attached file, in R)? At least by a row line - instead of seeing just a black dot.

 

Because now I need to press on every point to see which point is which line, and this is not at all convenient.


Thank you

4 REPLIES 4
txnelson
Super User

Re: Labelling data points in PCA Biplot by a characteristic

A simple way to do this, is to go to the column in the data table you want to be displayed, and to right click on the column and select "Label/UnLabel". 

pca1.PNG

Then when you hover over, or click on a data point, you will see the label.

pca2.PNG

If you want the labels to become persistent, all you need to do is to select the "Label/UnLabel" option for the rows in the data table you want the labels to be continuously displayed.

pca3.PNG 

pca4.PNG

Jim
dainius
Level II

Re: Labelling data points in PCA Biplot by a characteristic

Can I have this info drawn in the graph though - instead of the black points? 

I would be much more useful to me to see them all at once than to wander on each individual point.

txnelson
Super User

Re: Labelling data points in PCA Biplot by a characteristic

See my addition to my initial response.
Jim
dainius
Level II

Re: Labelling data points in PCA Biplot by a characteristic

Ok, maybe the question was unclear. This is what I would love to see (attached).

My samples are named S1, S2, S3, S4 and so on.

For every point, instead of a point, I would want to see S1, S2 or S3 written instead of a point. 

I have deleted some black points on your plot to show what I actually mean.

 

I think it makes it much easier to spot the groups of the samples and to get an idea of how are they scattered.

It's kinda a preliminary grouping. Now by an immediate looking I can see that S1 clusters together. Ok, let's investigate this further. Instead of manually pressing on 100 points to see which category they belong to.