Probability and Statistics The newest additions and improvements to probability and statistics functionality focus on data located in space and time. The new spatial analysis functions allow you to find the central location or central data element, depending on the distance function specific to the data type: geolocations, images, texts, dates, etc. In addition, more robust measures of location and dispersion were added to provide better analysis for numeric data with outliers and coming from heavy-tail distributions. Numerous performance improvements allow for enhanced data examination, processing and modeling. New robust location measure (spatial median) supporting numeric and geodetic data. » New location measure (central feature) working with many types of data. » New robust location measures for numeric data with outliers. » New robust dispersion measures for numeric data with outliers. » New heavy-tailed Pareto–Pickands distribution to model exceedance. » New Hjorth distribution with bathtub-shaped hazard function for reliability modeling. » Extended hypothesis tests of dispersion. » Estimation support for more matrix distributions. » Faster computation of some descriptive statistics. » Faster simulation and estimation of parametric distributions. » Faster computations with nonparametric distributions. » Improved processing of time series with calendar time stamps. » Related Examples Compute Urban Population Spatial Center » Spatial Statistics on the Surface of the Earth » Compute Central Location with Different Distance Functions » Central Feature of Texts and Images » When Mean Is Useless – Robust Location Measures » Advantage of Robust Location Measures for Heavy-Tailed Distributions » Robust Location Measures for Heavy-Tailed Data » Robust Dispersion Measures in the Presence of Outliers » Fitting Power-Tail Data with Pareto-Pickands Distribution » Modeling Floods with Pareto-Pickands Distribution » Modeling Bathtub Failure Rates with Hjorth Distribution » Extended Hypothesis Tests for Dispersion Equivalence »
