fastmath.kernel — generateme/fastmath 3.0.0-alpha1

Various kernel functions.

• RBF (double -> double functions)
• vector kernels (vector x vector -> double function; may be positive definite, conditional positive definite, positive semi-definite, mercer)
• density estimation
• some kernel operations

bandwidth clj

(bandwidth kernel data h)

Returns infered bandwidth (h).

h can be one of:

• :nrd - rule-of-thumb (scale=1.06)
• :nrd0 - rule-of-thumb (scake=0.9)
• :nrd-adjust - kernel specific adjustment of :nrd, doesn't work for silverman and cauchy
• :rlcv - robust likelihood cross-validation
• :lcv - likelihood cross-validation
• :lscv - least squares cross-validation

exp clj

(exp k)

(exp k t)

Kernel wraper. exp of kernel k with optional scaling value t.

kernel clj multimethod

Create vector kernel.

Vector kernel returns a number for two vectors (or numbers)

Kernels can be Mercer, positive definite, conditional positive definite, positive semi-definite or other.

kernel-density clj

(kernel-density kernel data)

(kernel-density kernel data bandwidth)

(kernel-density kernel data bandwidth info?)

Returns kernel density estimation function, 1d

Arguments:

• kernel - kernel name or kernel function
• data - data
• params - a map containing:
  • :bandwidth - bandwidth h
  • :binned? - if data should be binned, if true the width of the bin is bandwidth divided by 5, if is a number then it will be used as denominator. Default: false.

:bandwidth can be a number or one of:

• :nrd - rule-of-thumb (scale=1.06)
• :nrd0 - rule-of-thumb (scake=0.9)
• :nrd-adjust - kernel specific adjustment of :nrd, doesn't work for silverman and cauchy
• :rlcv - robust likelihood cross-validation
• :lcv - likelihood cross-validation
• :lscv - least squares cross-validation

kernel-density-ci clj

(kernel-density-ci kernel data)

(kernel-density-ci kernel data bandwidth-or-params)

Create function which returns confidence intervals for given kde method.

Check 6.1.5 http://sfb649.wiwi.hu-berlin.de/fedc_homepage/xplore/tutorials/xlghtmlnode33.html

Arguments:

• kernel - kernel name
• data - sequence of data values
• params - map with other parameters (including kernel-density parameters)
  • alpha - confidence level, default: 0.05

Returns three values: density, lower confidence value and upper confidence value

kernel-list clj

mult clj

(mult k1)

(mult k1 k2)

(mult k1 k2 k3)

(mult k1 k2 k3 & r)

Kernel wrapper. Multiply two or more kernels.

rbf clj multimethod

RBF kernel creator. RBF is double->double function.

Refer fastmath.kernel.rbf namespace for details.

scale clj

(scale k scale)

Kernel wrapper. Scale kernel result.

wmean clj

(wmean kernels)

(wmean kernels weights)

Kernel wrapper. (Weighted) mean of kernel results.