Isatis latest release helps you further in search neighborhood definition

April 25, 2016

Make informed decisions when defining kriging neighborhoods and provide evidence to support them. The new Neighborhood Statistics functionality in Isatis 2016 computes all sorts of statistics to help in the choice of relevant neighborhood parameters.

One cannot expect quality estimation without setting a correctly defined search neighborhood. But defining a search neighborhood well adapted to the geological and grade data may prove tedious and complex since many parameters may come into play. In particular, search neighborhood must be appropriate to grade anisotropies, sampling plan, sample quantity, etc.. Thus the need for a tool which gives maximum flexibility in neighborhood design.

Geovariances has always endeavored to provide Isatis’ users with the very best in geostatistics. Isatis Neighborhood Definition application is a good example as it has continuously been improved over the years, in our concern to respond users’ requests. New parameters have been introduced helping users to define the best neighborhood for their data. Among them, the number of samples which should be selected per line and per sector, capping outliers on the fly, etc..

Besides, new functionalities have delivered outputs to help in assessing estimation quality: Kriging Efficiency which measures the expected error for each block grade or Neighborhood Statistics in this latest release.

With the new Neighborhood Statistics, you may collect all sorts of statistics about neighbors and store them in variables that you can use later on for display, research, selection, etc.. They include the quantiles and statistics on neighbor values, on the distances between the targets and the selected neighbors, the number of selected neighbors per sector or of empty sectors, the number of active boreholes in the neighborhood, all of them can be stored in macrovariables, etc..

Neighborhood statistics comes in addition to the powerful interactive Test Window, available from any interpolation application. The Test Window allows you to focus the control where the defined neighborhood is not relevant and help you understand why.

We have put this expertise in Minestis built-in Kriging Neighborhood Analysis which simplifies the control procedure. It allows quickly evaluating the quality of a neighborhood in order to select the most adequate one regarding the actual input dataset and variography.