Tag: formula

Detect Outlier using Boxplot in Java

import org.apache.commons.math3.stat.descriptive.DescriptiveStatistics;
public boolean isOutlier(String itemId, String value) {

    if (monitoringItem.getProperties().getBoxPlotSize() > 0) {
        BoxPlot boxPlot = Statics.boxPlotRepository.findByItemId(itemId);

        DescriptiveStatistics descriptiveStatistics = new DescriptiveStatistics();
        double dValue = Double.parseDouble(value);

        for (ItemBoxPlot bItem : boxPlot.getBoxPlotList()) {
            double d = Double.parseDouble(bItem.getValue());

            descriptiveStatistics.addValue(d);
        }

        double Q1 = descriptiveStatistics.getPercentile(25);
        double Q3 = descriptiveStatistics.getPercentile(75);
        double IQR = Q3 - Q1;

        double highRange = Q3 + 3 * IQR;
        double lowRange = Q1 - 3 * IQR;

        if (dValue > highRange || dValue < lowRange) {
            return true;
        }
    }

    return false;
}

References
https://pupli.net/2019/05/31/detect-outlier-using-boxplot/

Detect Outlier using Boxplot

Order the data from least to greatest then calculate these values:

  • Q1 – quartile 1, the median of the lower half of the data set
  • Q2 – quartile 2, the median of the entire data set
  • Q3 – quartile 3, the median of the upper half of the data set
  • IQR – interquartile range, the difference from Q3 to Q1
  • Extreme Values – the smallest and largest values in a data set

Outliers

In order to be an outlier, the data value must be:

  • larger than Q3 by at least 1.5 (some times 3) times the interquartile range (IQR), or
  • smaller than Q1 by at least 1.5 (some times 3) times the IQR.

References
https://en.wikipedia.org/wiki/Box_plot
https://www.shmoop.com/basic-statistics-probability/box-whisker-plots.html
https://pupli.net/2019/05/31/detect-outlier-using-boxplot-in-java/