AREAXTAB

Description

AREAXTAB generates an area cross-tabulation report containing correlation coefficients (Chi Square coefficients). These coefficients gauge the extent of correlation between two raster layers.

Parameters

Name	Type	Length	Value range
Input1: Input raster channel for rows *	Raster port	1 - 1
Input2: Input raster channel for columns *	Raster port	1 - 1
Bitmap: Input bitmap	Bitmap port	0 - 1
Units	String	0 - 1	Square Meters, Square Kilometers, Hectares, Square Miles, Acres, Square Feet Default: Hectares
Report	String	0 - 192	See parameter description

* Required parameter

Parameter descriptions

Input1: Input raster channel for rows

Specifies the raster data channel to use in generating part of the report. This data is used in the table rows. The raster can be of any data type.

Input2: Input raster channel for columns

Specifies the raster data channel to use in generating part of the report. This data is used in the table columns. The raster can be of any data type.

Bitmap: Input bitmap

If specified, the report is generated only for pixels under the bitmap.

Units

Specifies the area units to use in the report.

The default value is Hectares.

Report

Specifies where to direct the generated report.

Available options are:

• LOG: generates a report on the terminal (default)
• <modulename.RPT>: appends a report to the module's RPT file
• DISK: generates a report on file "IMPRPT.LST"

Details

AREAXTAB generates an ASCII report that enumerates each pixel value along with its name, the area percentage, the row percentage, and the column percentage covered by the pixel value.

Optionally, the report can be generated for only the pixel values that fall under a specified bitmap.

Pixels with a NO_DATA_VALUE are omitted from the report.

Chi Square coefficients measure the degree of correlation, association, or dependencies of a sample. Because the contingency table varies in dimension, three different estimates are generated, as described below.

Contingency Coefficient

The Contingency Coefficient (C) theoretically lies between 0.0 and 1.0, but does not always reach 1.0, even though the rasters seem completely associated.

Tschuprow's T

Tschuprow's T varies between 0.0 (for independence) and 1.0, but can only attain the maximum value in square tables.

Cramer's V

Cramer's V corrects some of the defiencies of C and T in that it achieves its maximum in assymmetic arrays. The V is always at least as large as T.

References

Reynolds, H. T., The Analysis of Cross-Classifications, New York, The Free Press, 236 pages.