Descriptive statistics, as follows from its name, describes a distribution by means of figures which characterize distribution parameters. These parameters can be divided into three main groups: central tendency measures, dispersion measures and shape measures.

Central tendency measures

Central tendency measures characterize the central value around which random variable values are distributed. These measures are the mean (first moment of distribution) and the median. The mean is used for the description of close to normal distributions. If the distribution essentially differs from normal (for example, if it has long and wide tails), the mean will converge to "true" mean too slowly. In that case, it is reasonable to use the median to estimate central value.

Dispersion measures

Dispersion measures characterize scattering whose random value is distributed around its central value. These measures are dispersion (and related with it standard deviation) and average deviation. As for central tendency measures, one of these measures (dispersion) is good for close to normal distributions while the second one (average deviation) is better for long-tailed distributions.

Shape measures

Shape measures describe common characteristics of distribution appearance: skewness, kurtosis, percentile.

Report a bug

C#

C# 1.0 source.
descriptivestatistics.csharp.zip - Descriptive statistics

C++

C++ source.
descriptivestatistics.cpp.zip - Descriptive statistics
ablas.zip - optimized basic linear algebra subroutines with SSE2 support (for C++ sources only)

Delphi

Delphi source.
Can be compiled under FPC (in Delphi compatibility mode).
descriptivestatistics.delphi.zip - Descriptive statistics

Visual Basic 6

Visual Basic 6 source.
descriptivestatistics.vb6.zip - Descriptive statistics

Zonnon ^beta

Zonnon source.
Zonnon is an experimental language developed at ETH Zurich.
See www.zonnon.ethz.ch for more information.
descriptivestatistics.zonnon.zip - Descriptive statistics