CHistogram:SetCorrelated
The SetCorrelated method informs Mira that the data are correlated so that an alternate, and faster, histogramming strategy will be chosen for the Calc method. Correlation means that data values at similar array index have a value that is more similar than the range of values being used for the histogram. At the two extremes of correlation, date that are sorted into ascending or descending order would provide the maximum speed gain while data in completely random order would offer no speed gain. By setting this option to true, the speed gain can be as great as 0.58 the computation time. If you do not use this method, the data are considered uncorrelated.
CHistogram:SetCorrelated( bTrue )
The speed gain offered for correlated data has been evaluated using random data and sorted data. The speed increase is given in the following table.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|