Child pages
  • Estimation - Inverse Distance
Skip to end of metadata
Go to start of metadata

Inverse Distance

Calculated by taking weighted average of the value of the point to be estimated and composite data to be used in estimating this point. Weights are calculated inversely proportional to the distance between the point to be estimated and composite data. This proportion degree is determined with alpha parameter. Alpha can take values between 1 and k. The more the value of this parameter, the closer it gets to nearest adjacent method. 

"Attribute" should be chosen from the window (variable to be estimated), which is opened by entering Reverse Distance estimation operation. 

If the average values of blocks will be estimated, number of block divisions should be entered for x, y and z directions. Block division number should be 1 for x, y and z directions in point estimation. Division numbers of block estimations are dependent on the anisotropy direction and degree of the variable to be estimated. These numbers can be taken as equal to three in isotropy. 

After accessing the Reverse Distance estimation operation, parameters relating to estimation geometry must be entered. These are alpha, distance, maximum number of samples. Alpha value can practically be taken as 1 or 2.  

Maximum Number of Neighbors is the maximum number of samples to be used in an estimation of a point or block. 

Distance means the radius of the sphere from which the samples shall be chosen. If the variable to be estimated is anisotropic, samples should be chosen from inside of an ellipsoid instead of a sphere with a "distance" radius. The geometry of the ellipsoid (axis lengths (radius) and angles) can be determined with variogram analysis. 

If you want to keep the estimation results, you must determine the file and index of the file by clicking three dot (...) button under "Estimation Result File". The file is saved in *.sgems extension. 

 

 

 Video
 

  • No labels