pyrimidine.benchmarks subpackage

pyrimidine.benchmarks.approximation module

class pyrimidine.benchmarks.approximation.Function1DApproximation(function, lb=0, ub=1, basis=[<function <lambda>>, <function <lambda>>, <function <lambda>>, <ufunc 'sin'>, <ufunc 'cos'>, <ufunc 'tan'>, <ufunc 'exp'>, <function <lambda>>, <function <lambda>>])

Bases: BaseProblem

pyrimidine.benchmarks.approximation.lin_comb(x, coefs, basis)

pyrimidine.benchmarks.cluster module

class pyrimidine.benchmarks.cluster.KMeans(X, n_components=2)

Bases: BaseProblem

KMeans clustering Problem

ERM: min J(c,mu) = sum_c sum_{x:c} ||x-mu_c||

static random(N, p=2)

pyrimidine.benchmarks.fitting module

pyrimidine.benchmarks.linear_model module

pyrimidine.benchmarks.linear_model.fun(x)
pyrimidine.benchmarks.linear_model.lsq(X, A, B, alpha=0.1)

pyrimidine.benchmarks.matrix module

class pyrimidine.benchmarks.matrix.NMF(M)

Bases: BaseProblem

static random(N=500, p=100)

pyrimidine.benchmarks.neural_network module

class pyrimidine.benchmarks.neural_network.MLP(X, Y)

Bases: BaseProblem

Y = A2f(A1X+b1)+b2

static random(N=100, p=2)
class pyrimidine.benchmarks.neural_network.RNN(X, Y)

Bases: BaseProblem

Yt+1 = A2f(A1Xt+C1Zt+b1)+b2 Zt+1 = C2g(A1X+C1Zt+b1)+c2

static random(N=100, p=2)

pyrimidine.benchmarks.optimization module

class pyrimidine.benchmarks.optimization.CurvePath(x, y)

Bases: ShortestPath

class pyrimidine.benchmarks.optimization.FacilityLayout(F, D)

Bases: BaseProblem

F: F D: D

static random(self, n)
class pyrimidine.benchmarks.optimization.Knapsack(w, c, W=0.7, M=100)

Bases: BaseProblem

Knapsack Problem

max sum_i ci xi s.t. sum_i wi xi <= W xi = 0 / 1 (choice variable) where ci is in c, wi is the coresponding weight of ci

argsort()
static example(W=0.7)
property n_bags
static random(n_bags=50, W=0.7)
property sorted
class pyrimidine.benchmarks.optimization.MLE(pdf, x)

Bases: BaseProblem

static random(size=300)
class pyrimidine.benchmarks.optimization.MinSpanningTree(nodes, edges=[], weights=None)

Bases: BaseProblem

class pyrimidine.benchmarks.optimization.MixMLE(pdfs, x)

Bases: BaseProblem

logpdf(x, t, a)
static random(n_observants=300, n_components=2)
class pyrimidine.benchmarks.optimization.MultiKnapsack(ws, cs, W=0.7, M=100)

Bases: BaseProblem

Multi Choice Knapsack Problem

max sum_ij cij xij s.t. sum_ij wij xij <= W {xij} is where xij is the choice variable, that is sum_j xij=1

argsort()
static random(size=(4, 5, 2, 6, 5, 7, 8, 3), W=0.7)
class pyrimidine.benchmarks.optimization.ShortestPath(points)

Bases: BaseProblem

static random(N)

pyrimidine.benchmarks.others module

class pyrimidine.benchmarks.others.Kantorovich(a=0.5, b=1)

Bases: BaseProblem

a: a b: b

pyrimidine.benchmarks.special module

Spcical functions for benchmark

pyrimidine.benchmarks.special.alpine(x)
pyrimidine.benchmarks.special.griewangk(n=5)
pyrimidine.benchmarks.special.hansen(n=5)
pyrimidine.benchmarks.special.michalewiez(n=5)
pyrimidine.benchmarks.special.rastrigrin(x: ndarray)
pyrimidine.benchmarks.special.rosenbrock(x: ndarray)
pyrimidine.benchmarks.special.schaffer(x: ndarray)