PrincompGUI {DASplusR}R Documentation

GUI for Principal Component Analysis

Description

Starts the graphical user interface for principal component analysis. When the OK button is clicked, princompDAS is executed and a summary of the results is shown in the output window of DAS+R.

Usage

## DAS+R GUI:
Advanced --> Princ. Components Anal. --> Compute PCs

## R Function Usage:
PrincompGUI()

Dialog Elements

Select Variables:
If at least two variables are preselected in DAS+R, only these are shown in the listbox. Otherwise, it contains all numerical variables of the active data set. The radiobuttons on the right can be used to select or deselect all variables.
Log-Transformation:
Check this checkbox for using log-transformation. A dialog for selecting the variables to be transformed will pop up when the OK button is clicked.
Previously Calculated Principal Components:
Checking this checkbox allows the selection of a previously calculated object from the drop-down list on the right. The drop-down list contains objects of class "princomp". Such an object is only listed here if the function call used for its computation satisfies one of the following two conditions: either the active data set has been used as the data argument, or no data argument has been supplied. Because an already existing object is used and no calculation will be done, all other dialog elements, except the checkboxes for the scree plot and the biplot, are disabled. They cannot be used until this checkbox is unchecked again.
Correlation (Standardization):
This checkbox determines whether the principal components will be calculated using the correlation matrix or the covariance matrix. The correlation matrix cannot be used if there are any constant variables.
Robust Covariance Matrix Estimation:
If this checkbox is checked, a robust estimation method for the covariance matrix can be selected from the drop-down list on the right. Possible choices are MCD and OGK estimation. Also, the checkbox for using a previously calculated covariance matrix is disabled. It can be used again when this checkbox is unchecked again.
Previously Calculated Covariance Matrix:
By checking this checkbox, a previously calculated covariance matrix to be used in PCA can be selected. The drop-down list on the right shows all matrices and covariance lists, i.e., lists that contain elements named center and cov, as returned by cov.wt (or covMcd from package robustbase). Scores cannot be computed if the selected object is a matrix, as no estimate of the central location of the data is supplied. In any case, make sure that the same variables are selected that have been used to calculate the covariance matrix. Additionally, log-transformation and robust estimation of the covariance matrix cannot be used anymore. Hence the corresponding dialog elements are disabled until this checkbox is unchecked again.
Plots:
Check the checkboxes for the scree plot and/or the biplot to produce the respective plots. The corresponding dialogs are called when the OK button is clicked.
Name for the Object to be Returned:
The combo box can be used to enter a name, or to select the name of an already existing object. Note that an existing object will be overwritten without asking for confirmation. The drop-down list contains objects of class "princomp". One of the following two conditions on the function call used to compute such an object has to hold: either the active data set has been used as the data argument, or no data argument has been supplied. Otherwise, the object is not shown in the list.
Default:
Resets all the dialog elements to their default values.
OK:
Closes the dialog and executes princompDAS. A summary of the results is shown in the output window of DAS+R.
Cancel:
Closes the dialog without executing princompDAS.
Help:
Shows this help file.

Author(s)

Andreas Alfons <andreas.alfons@student.tuwien.ac.at>

References

Mardia, K.V., Kent, J.T. and Bibby, J.M. (1979) Multivariate Analysis. Academic Press.

Johnson, R.A. and Wichern, D.W. (2002) Applied Multivariate Statistical Analysis. Prentice Hall. 5th edition.

Reimann, C., Filzmoser, P., Garrett, R.G. and Dutter, R. (2008) Statistical Data Analysis Explained. Wiley & Sons.

See Also

princompDAS, LogTransform, ScreeplotGUI, BiplotGUI.


[Package DASplusR version 0.0-2 Index]