Statistics

The Function Development Kit (FDK) provides access to many statistical built-in functionalities.

The type 'Value' provides descriptive functions for selections, table cells, attributes, ...

Value

Minimum

Maximum

Range

Arithmetic mean

Geometric mean

Harmonic mean

Quadratic mean

Median

Quartile

Decile

Percentile

Sum

Product

Sum of squares

Sum of logs

Variance

Standard deviation

Skewness

Kurtosis

Three correlation coefficents are supported: Pearson, Spearman and Kendall

Pearson's Correlation

Spearman's Correlation

Kendall's Correlation

The covariance of two variables within selections can be computed quite easily:

Covariance

Simple linear regression with direct observations:

Simple linear regression

Simple linear regression on selections:

Simple linear regression on selections

Multiple linear regression with direct observations:

Multiple linear regression

Multiple linear regression on selections:

Multiple linear regression on selections

Many common statistical tests are provided by built-on functions.
They can either be performed directly via type 'InferenceRoot' or on selections via type 'SelectionInference'.

InferenceRoot SelectionInference


T-test with direct observations:

T-test

T-test on selections:

T-test on selections

Binomial test:

Binomial test

Supported tests:

• T
• Paired T
• Homoscedastic T
• Chi-Square
• G
• Binomial
• One-way ANOVA
• Kolmogorov–Smirnov
• Mann–Whitney U
• Wilcoxon signed-rank

Many common distributions can be created via the type 'DistributionRoot'.

DistributionRoot

Create a distribution

Supported distributions:

• Beta
• Binomial
• Cauchy
• Chi squared
• Constant
• Enumerated
• Exponential
• F
• Gamma
• Geometric
• Gumbel
• Hypergeometric
• Laplace
• Levy
• Logistic
• Log normal
• Nakagami
• Normal
• Pareto
• Pascal
• Poisson
• T
• Triangular
• Uniform
• Weibull
• Zipf