Index
Single Raster Analysis Tools IV Computing Slope and Aspect with SURFACE	A-Z	Single Raster Analysis Tools VI Interpolation Methods in IDRISI

Single Raster Analysis Tools V

• QUADRAT and CENTER - 2 modules for point pattern analysis
  Epidemiological, geological, ecological ... questions sometimes require a means of describing point distribution patterns. In the Analysis/Statistics submenue of IDRISI you will find 2 modules, that may give you some assistance - QUADRAT and CENTER. They assume the cellvalues of the input image as equivalents to the count of points (occurrences) that fall within that cells (additionally it can be a weighting factor for each cell in CENTER).
  In both modules IDRISI results with statistical summary informations written out to a textedit window
  • CENTER can be used e.g. to show the shifting in population density over time by repeating the calculation for time series images.

  results of CENTER
  1.	Image analyzed : popden	name of analyzed file
  2.	Mean center falls in col 250 / row 395	the col/row position of the weighted mean center
  3.	Standard radius in cell widths is 36716	analogue to standard deviation in univariate distributions; tells expresses the dispersion of points about the mean center
  4.	X coordinate of the mean center in M : 253899.688721	same as col in 2., but ref. units
  5.	Y coordinate of the mean center in M : -679978.179932	same as row in 2., but ref. units
  6.	Standard radius on the ground expressed in M : 183577680.000000	same as cells in 3., but ref. units
  7.	Coefficient of Relative Dispersion (%) : 14268.986511 (= CRD)

  • QUADRAT computes mean, variance, variance/mean ratio, t-test result, significance level and an estimation of the "ideal" quadrat size, as the results of the variance/mean ratio can be biased by that size. Ratios arround 1.0 indicate a random point pattern, ratios < 1.0 suggest distributed patterns and ratios > 1.0 clustered ones. But the analyst has to pay attention to possible autocorrelative effects (AUTOCORR) that may lead to distortions in the results. QUADRAT allows the declaration of a mask image (1's where cells should be computed, 0's where should not = BOOL image).

	Index
Single Raster Analysis Tools IV Computing Slope and Aspect with SURFACE	A-Z	Single Raster Analysis Tools VI Interpolation Methods in IDRISI

last modified: | Comments to Eric J. LORUP