In ILWIS, there are four Statistical operations which calculate a matrix:

• covariance matrix by the Variance-Covariance matrix operation;
• correlation matrix by the Correlation matrix operation;
• covariance matrix and matrix of eigenvectors by the Principal Component operation;
• correlation matrix and matrix of eigenvectors by the Factor Analysis operation.

All these operations can be found on the Operations menu, Statistics, Map List. As the matrices are the result of an operation, the matrices cannot be edited.

Furthermore,

• you can calculate a confusion matrix after an image classification. For more information, see Table window : Confusion matrix or How to calculate a confusion matrix.
• you can show an output cross table as a matrix when the input maps of the Cross operation used a class or ID domain. For more information, see Cross : functionality.

Names of matrices:

When performing the Variance-Covariance operation or the Correlation operation, the matrix is not stored as a separate object; the matrix is stored in the input map list.

When performing the Principal Component or Factor analysis operation, you can specify a name for the matrix yourself; the output map list which will contain the transformed output maps will obtain the same name as the matrix.

In ILWIS 3, object names comply with Windows long file names. Also Universal Naming Convention (UNC) paths are supported. For more information, see How to use long object names.

To display a matrix:

To display a covariance matrix, perform the Variance-Covariance operation.

To display a correlation matrix, perform the Correlation operation.

To display the output matrices of the Principal Component and Factor Analysis operations:

• click the Show button in the dialog box of the Principal Component and Factor Analysis operation, or
• double-click a matrix in a Catalog, or
• use the right mouse button on a matrix in a Catalog, and select Open from the context-sensitive menu, or
• from the File menu in the Main window, choose Open and select a matrix.

The matrix will be calculated as soon as it is shown for the first time in a matrix window. The matrix window also displays some Additional Info.

Technical information:

A matrix consists of an ASCII object definition file (.MAT).

By viewing the properties of a matrix, you can see on which map list the matrix depends, etc. When the matrix has been calculated, you can also view Additional Info. For more information, see the Matrix Properties sheet.

Variance-covariance matrix:

Shows the variance within each map/band and the covariance between all map/band pairs in the input map list. The mean and standard deviation of each input map are also shown. In the properties of the input map list, the Optimum Index Factors (OIF) are listed.

Correlation matrix:

Shows the correlation between all map/band pairs in the input map list. The mean and standard deviation of each input map are also shown. In the properties of the input map list the Optimum Index Factors (OIF) are listed.

Principal component matrix:

Stores eigenvectors for the new output maps (axes) and the percentages of total variance contributed by each eigenvalue. The operation creates (defines) an output matrix and an output map list simultaneously. When the output map list is opened (calculated), the output raster maps are created and calculated.

Factor analysis matrix:

Stores eigenvectors for the new output maps (axes) and the percentages of total variance contributed by each eigenvalue. The operation creates (defines) an output matrix and an output map list simultaneously. When the output map list is opened (calculated), the output raster maps are created and calculated.

See also:

Variance-Covariance matrix

Correlation matrix

Principal Component Analysis

Factor analysis

How to calculate Optimum Index Factor (OIF)

Table window : Confusion matrix

How to calculate a confusion matrix