Angara.Statistics


Statistics

Namespace: Angara

Nested types and modules

TypeDescription
Distribution
MT19937
qsummaryType
summaryType
ModuleDescription
Serialization

Functions and values

Function or valueDescription
correlation x y
Signature: x:float [] -> y:float [] -> float
dct rxs
Signature: rxs:float [] -> float []

Descrete cosine transform.

draw gen d
Signature: gen:MT19937 -> d:Distribution -> float
e
Signature: float<MeasureOne>

natural logarithm base

fft xs
Signature: xs:Complex [] -> Complex []

Fast Fourier transform.

histogram_ n xmin xmax xs
Signature: n:int -> xmin:float -> xmax:float -> xs:seq<float> -> int []
idct rxs
Signature: rxs:float [] -> float []

Inverse discrete cosine transform.

ifft xs
Signature: xs:Complex [] -> Complex []

Inverse Fast Fourier Transform.

improbable
Signature: float<MeasureOne>

The smallest positive normalized float value

kde n0 xs
Signature: n0:int -> xs:seq<float> -> float [] * float []

Gaussian kernel density estimator for one-dimensional data, using the method of Botev et al.

kde2 n0 min max sample
Signature: n0:int -> min:float -> max:float -> sample:seq<float> -> float [] * float []

Gaussian kernel density estimator for one-dimensional data, using the method of Botev et al.

log_improbable
Signature: float<MeasureOne>

Logarithm of improbable

log_pdf d v
Signature: d:Distribution -> v:float -> float<MeasureOne>

Logarithm of a Probability Distribution Function

log_tolerance
Signature: float<MeasureOne>

Logarithm of tolerance

log2pi
Signature: float<MeasureOne>

1/2 * log 2π

logistic x
Signature: x:float<MeasureOne> -> float<MeasureOne>

Sigmoidal function that maps [-infinity,infinity] interval onto [0,1]

logit p
Signature: p:float<MeasureOne> -> float

Inverse logistic transform

maxint
Signature: float<MeasureOne>

Maximum exact integer maxint+1.0 = maxint && maxint-1.0 < maxint

pi
Signature: float<MeasureOne>

π

pi2
Signature: float<MeasureOne>

qsummary data
Signature: data:seq<float> -> qsummaryType

Produces quantile summary of the data.

ridders tolerance (lb, ub) f
Signature: tolerance:float -> (lb:float * ub:float) -> f:(float -> float) -> float option

Root of a function using Ridders method.

sqrt2pi
Signature: float<MeasureOne>

sqrt 2π

summary data
Signature: data:seq<float> -> summaryType

Produces cumulant summary of the data using fast one-pass algorithm.

tolerance
Signature: float<MeasureOne>

1.0 - tolerance < 1.0 && 1.0 - 0.5*tolerance = 1.0

within ulps a b
Signature: ulps:uint32 -> a:float -> b:float -> bool

Approximate comparison of two double values. Tolerance ulps is in units of least precision.

Fork me on GitHub