kixi.stats.distribution — kixi/stats 0.4.4

bernoulli clj/s

(bernoulli p)

Returns a Bernoulli distribution. Params: p ∈ [0 1]

beta clj/s

(beta {:keys [alpha beta] :or {alpha 1.0 beta 1.0}})

Returns a beta distribution. Params: {:alpha ∈ ℝ, :beta ∈ ℝ}

beta-binomial clj/s

(beta-binomial n {:keys [alpha beta] :or {alpha 1.0 beta 1.0}})

Returns a beta distribution. Params: n ∈ ℕ, {:alpha ∈ ℝ, :beta ∈ ℝ}

binomial clj/s

(binomial {:keys [n p]})

Return a binomial distribution. Params: {:n ∈ ℕ, :p ∈ [0 1]}

categorical clj/s

(categorical ks ps)

Returns a categorical distribution. Params: [k1, ..., kn], [p1, ..., pn] where k1...kn are the categories and p1...pn are probabilities. Probabilities should be >= 0 and sum to 1

cdf clj/s

chi-squared clj/s

(chi-squared k)

Returns a chi-squared distribution. Params: k ∈ ℕ > 0

dirichlet clj/s

(dirichlet as)

Returns a Dirichlet distribution. Params: [a1...an] ∈ ℝ >= 0

dirichlet-multinomial clj/s

(dirichlet-multinomial n as)

Returns a Dirichlet-multinomial distribution. Params: n ∈ ℕ, [a1...an] ∈ ℝ >= 0

draw clj/s

(draw distribution)

(draw distribution {:keys [seed]})

Returns a single variate from the distribution. An optional seed long will ensure deterministic results

exponential clj/s

(exponential rate)

Returns an exponential distribution. Params: rate ∈ ℝ > 0

f clj/s

(f d1 d2)

Returns an F distribution. Params: d1 ∈ ℕ > 0, d2 ∈ ℕ > 0

gamma clj/s

(gamma {:keys [shape scale] :or {shape 1.0 scale 1.0}})

Returns a gamma distribution. Params: {:shape ∈ ℝ, :scale ∈ ℝ}

iqr clj/s

(iqr distribution)

Returns the interquartile range

maximum clj/s

median clj/s

(median distribution)

Returns the median

minimum clj/s

multinomial clj/s

(multinomial n ps)

Returns a multinomial distribution. Params: n ∈ ℕ > 0, [p1, ..., pn] where p1...pn are probabilities. Probabilities should be >= 0 and sum to 1

normal clj/s

(normal {:keys [mu sd]})

Returns a normal distribution. Params: {:mu ∈ ℝ, :sd ∈ ℝ}

poisson clj/s

(poisson lambda)

Returns a Poisson distribution. Params: lambda ∈ ℝ > 0

quantile clj/s

sample clj/s

(sample n distribution)

(sample n distribution {:keys [seed]})

Returns n variates from the distribution. An optional seed long will ensure deterministic results

sample-summary clj/s

(sample-summary n distribution)

(sample-summary n distribution {:keys [seed]})

Returns a summary count of each variate for a sample of a given length from a discrete distribution such as the Bernoulli, binomial or categorical. An optional seed long will ensure deterministic results

summary clj/s

(summary distribution)

Returns the 5-number distribution summary and the interquartile range.

uniform clj/s

(uniform a b)

Returns a uniform distribution. Params: a ∈ ℝ, b ∈ ℝ

weibull clj/s

(weibull {:keys [shape scale] :or {shape 1.0 scale 1.0}})

Returns a weibull distribution. Params: {:shape ∈ ℝ >= 0, :scale ∈ ℝ >= 0}