Pitney Bowes MapInfo Vertical Mapper - Manual (Page 165)

Chapter 9: Data Analysis
User Guide
163
• Anisotropy view diagram shows the directional trends in the data. A wide ellipse indicates that
there is a greater degree of correlation of the variances between sample pairs in that direction.
Conversely, a narrow ellipse indicates that there is a smaller degree of correlation. The angle
setting for each model determines the degree of rotation for each ellipse.
• The Anistropic Modeling check box enables you to build more than one model curve for the
different directions analyzed. Once you enable this check box, the Angle and Anisotropy settings
become available for each chosen variogram model.
• Suggested Model analyzes the experimental variogram and chooses the variogram model that
best represents it. In some cases, it may not be possible to automatically generate a model. In
this case, a warning message appears and you must set a model manually.
The Nugget box enables you to force the semivariogram to pass through the y-axis at a higher
value. This has a smoothing effect on the kriging process and prevents it from acting as an exact
interpolation in scattered areas of high concentration.
Spatial Correlation
Spatial correlation analysis investigates relationships between grids. Because correlation analysis is
a statistical technique involving calculation, only numerical grids can be analyzed.
Correlation is a measure of correspondence and is expressed as a coefficient, which is a value
between -1 and 1. For positive values, the higher the value, the more closely the grids correspond.
For example, a value of 0.85 shows a relatively high correlation, whereas a value of 0.2 shows
relatively weak correlation. Negative values show an inverse correlation between grids (for example,
if there is strong negative correlation, when the values in one grid are low, the values in another are
correspondingly high). For example, a value of -0.85 shows relatively strong negative correlation,
whereas a value of -0.2 shows relatively weak correlation.
Correlation Matrix
Using a correlation matrix, you can examine a selection of numeric grids for correlation. All
permutations of pairs of grids in the selection are analyzed, and the results are returned in a matrix.
The correlation coefficient is displayed for each pair of grids.