simulate_langevin_diffusion_exp_4: Exact Algorithm for langevin diffusion with pi =...

In rchan26/exp4Tempering: Hierarchical and Sequential Monte Carlo Fusion for exp(-(x^4)/2)

Description

Simulate langevin diffusion using the Exact Algorithm with pi = exp(-(beta*(x-mean)^4)/2)

Usage

simulate_langevin_diffusion_exp_4(x0, y, s, t, K, mean = 0, beta,
  accept_prob = FALSE)

Arguments

x0	start value
y	end value
s	start time
t	end time
K	lower bound of the phi function
mean	mean value
beta	real value between 0 and 1
accept_prob	boolean value: defaults to F, determined whether or not to return acceptance prob (T) or a skeleton of the diffusion (F)

Value

accept_prob real value if accept_prob = TRUE

layered_bb matrix holding skeleton of the diffusion

Examples

lower_bound <- optimise(function(x) phi_function_exp_4(x, beta = 1/2), interval = c(0, 100), maximum = FALSE)$objective
test <- simulate_langevin_diffusion_exp_4(x0 = 0, y = 0.3, s = 0, t = 1, K = lower_bound, beta = 1/2, accept_prob = FALSE)
plot(x = test['time',], y = test['X',], type = 'l')

rchan26/exp4Tempering documentation built on June 20, 2019, 10:07 p.m.