Gcluster: Difference between revisions

From Eigenvector Research Documentation Wiki
Jump to navigation Jump to search
imported>Jeremy
(Importing text file)
imported>Benjamin
No edit summary
 
(2 intermediate revisions by 2 users not shown)
Line 1: Line 1:
===Purpose===
===Purpose===


K-means and K-nearest neighbor cluster analysis with dendrograms.
Internal function used with cluster analysis in the [[Analysis_GUI|Analysis interface]].


===Synopsis===
===Synopsis===


:gcluster(data,''labels'')
:n/a


===Description===
===Description===


gclster(data) performs a cluster analysis on the data matrix data using K-means or K-nearest neighbor clustering and plots a dendrogram showing distances between the samples. gcluster is a graphical user interface that calls the function cluster. The user can choose cluster method (K-means or KNN), and data scaling options. PCA can also be used on the data with distances based on raw scores or on a Mahalanobis distance measure.
This function is no longer an entry point for the clustering interface. To access a graphical user interface for cluster analysis, see the [[Analysis_GUI|Analysis interface]].
 
 
As of PLS_Toolbox/Solo 8.3, The gcluster GUI was updated to allow the use of Manhattan distance instead of Euclidean distance  (default). Note the three checkboxes 'Use PCA', 'Use Mahalanobis Distance', and 'Use Manhattan Distance'.
 
 
[[Image:gcluster_gui_manhattan.png|800px| Cluster in the Analysis GUI]]


gclster(data,''labels'') plots on the dendrogram sample names contained in the matrix of text ''labels''. ''labels'' can be entered as a matrix where each row is a label in single quotes and each label has the same number of characters.


Note: Calling gclster with no inputs starts the graphical user interface (GUI) for this analysis method.  
To keep the use of PCA scores (instead of the raw data) and the use of Manhattan Distance mutually exclusive, the following behaviors have been added to the GUI:
:*If 'Use PCA' is checked:
::* 'Use Mahalaobis Distance' will be enabled (to allow the user to check it).
::* 'Use Manhattan Distance' will be unchecked.
:*If 'Use Manhattan Distance' is checked:
::* 'Use PCA will be unchecked'.
::* 'Use Mahalanobis Distance' will be both unchecked and disabled.


===See Also===
===See Also===


[[cluster]], [[simca]]
[[analysis]], [[analysis_GUI]], [[cluster]], [[pca]], [[simca]], [[manhattandist]]

Latest revision as of 11:38, 16 August 2017

Purpose

Internal function used with cluster analysis in the Analysis interface.

Synopsis

n/a

Description

This function is no longer an entry point for the clustering interface. To access a graphical user interface for cluster analysis, see the Analysis interface.


As of PLS_Toolbox/Solo 8.3, The gcluster GUI was updated to allow the use of Manhattan distance instead of Euclidean distance (default). Note the three checkboxes 'Use PCA', 'Use Mahalanobis Distance', and 'Use Manhattan Distance'.


Cluster in the Analysis GUI


To keep the use of PCA scores (instead of the raw data) and the use of Manhattan Distance mutually exclusive, the following behaviors have been added to the GUI:

  • If 'Use PCA' is checked:
  • 'Use Mahalaobis Distance' will be enabled (to allow the user to check it).
  • 'Use Manhattan Distance' will be unchecked.
  • If 'Use Manhattan Distance' is checked:
  • 'Use PCA will be unchecked'.
  • 'Use Mahalanobis Distance' will be both unchecked and disabled.

See Also

analysis, analysis_GUI, cluster, pca, simca, manhattandist

Retrieved from "https://www.wiki.eigenvector.com/index.php?title=Gcluster&oldid=1167"