Semantics statistics of selected objects

Print Previous page Top page Next page

ic_0237 The given mode displays statistics (count of objects, the maximum, minimum, middle value of the semantic characteristic and middle-square value) of the specified semantic characteristic or  Н coordinate for three-dimensional objects and classifies (breaks into groups - classes) selected objects by values of semantics or height. Dependence of semantics values or Н coordinates from number of selected objects and classification of objects is displayed on the histogram or the graph.

There are three methods of classification in the task: a method of equal intervals, quantiles method, a method standart (root-mean-square) deviations.

At using method of equal intervals the identical difference between the top and bottom border in each class is established, i.e. each class contains an identical range of values. All range of values is divided onto five classes and each class is displayed on the histogram by its own color.

At using method of quantiles each class contains approximately identical number of objects. In this method also division onto five classes is made. Objects with identical values of characteristics are included into one class.

At using method of standart deviation each class is defined depending on removal of its values from the average value received for all objects. For delimitation of classes it is made or consecutive addition of value of a root-mean-square deviation  to the average value, or subtraction from average value. This method displays objects according to their position above or below average value.

The mode save the histogram and the statistical data into Excel at pressing of Report Excel button. The file of patterns semansel.xlt is in Mapcomp.dot catalogue.

For writing down the objects, divided onto classes by one of methods, into a file of objects lists it is necessary to press Object samples button. Thus into an existing or new lists file  the lists of  objects with names selsemstat1 … selsemstatN (where N - a number of classes) will be added. Each class is entered into the separate list, at absence of objects in a class the list will not be generated.

In future the lists of objects will be processed in modes Lists of objects and Union / crossing of sets.