README.md

In DouglasMesquita/robit: FLAMES: Flexible Link functions with AsyMptotES

FLAMES

Acknowlodgment

The authors would like to thank CAPES, CNPq and FAPEMIG for partial financial support.

Overview

FLAMES: Flexible Link function with AssyMptotES is a package able to fit binary regression under several link functions and two possible assymptotes parameters, c and d, in the form

Installation

# Install from CRAN (when available)
install.packages("FLAMES")
# Or the development version from GitHub
# install.packages("devtools")
devtools::install_github("DouglasMesquita/FLAMES")

Usage

library(FLAMES)

library(FLAMES) will load mcmc_bin, a function to fit binary regression models. The user can choose for MCMC based on ARMS or Metropolis Hastings algorithms. Also it is possible to fit logit, probit, cauchit, cloglog, loglog and robit regressions.

##-- Seed
set.seed(123456)

##-- Data
n <- 1000
n_cov <- 1

##-- Covariates
X <- matrix(rnorm(n*n_cov), ncol = n_cov)

##-- Coefficients
betas <- c(0, -1)
XBeta <- cbind(1, X)%*%betas

##-- c parameter
c1 <- 0.20
d1 <- 0.95

type_data = "cloglog"

##-- p and y
p <- FLAMES:::inv_link(x = XBeta, type = type_data, df = df)*(d1-c1) + c1
y <- rbinom(n = n, size = 1, prob = p)

bd <- data.frame(y = y, X)

##-- MCMC
nsim <- 2000
burnin <- 10000
lag <- 10

f <- y ~ X
type <- type_data

##-- ARMS ~ 4 minutes (soon in c++)
out_arms <- mcmc_bin(data = bd, formula = f,
                     nsim = nsim, burnin = burnin, lag = lag,
                     type = type, sample_c = TRUE, sample_d = TRUE,
                     method = "ARMS")

##-- ARMS ~ 55 seconds (soon in c++)
out_met <- mcmc_bin(data = bd, formula = f,
                    nsim = nsim, burnin = burnin, lag = lag,
                    type = type, sample_c = TRUE, sample_d = TRUE,
                    method = "metropolis")

##-- GLM
out_glm <- glm(formula = f, data = bd, family = "binomial")

cbind(glm = coef(out_glm),
      arms = coef(out_arms),
      metropolis = coef(out_met),
      real = betas)
#>                   glm        arms  metropolis real
#> (Intercept)  1.005911  0.03550848  0.03918668    0
#> X           -1.071551 -1.58034100 -1.54308670   -1

par(mfrow = c(2, 4), mar = c(3, 4, 1, 1))
plot(out_arms, ask = F)
plot(out_met, ask = F)

summary(out_arms)
#> $Call
#> mcmc_bin(data = bd, formula = f, nsim = nsim, burnin = burnin, 
#>     lag = lag, type = type, sample_c = TRUE, sample_d = TRUE, 
#>     method = "ARMS")
#> 
#> $Coeficients
#>                    mean std_error   lower_95   upper_95
#> (Intercept)  0.03550848 0.1914609 -0.3369406  0.3833066
#> X           -1.58034100 0.5113920 -2.5337717 -0.7536014
#> 
#> $`Other parameters`
#>                  mean  std_error  lower_95  upper_95
#> c parameter 0.3120890 0.08972683 0.1215077 0.4628683
#> d parameter 0.9322188 0.02085929 0.8912744 0.9722881
#> 
#> $`Fit measures`
#>        DIC  -2*LPML     WAIC
#> 1 1041.917 1042.101 1041.917

summary(out_met)
#> $Call
#> mcmc_bin(data = bd, formula = f, nsim = nsim, burnin = burnin, 
#>     lag = lag, type = type, sample_c = TRUE, sample_d = TRUE, 
#>     method = "metropolis")
#> 
#> $Coeficients
#>                    mean std_error   lower_95   upper_95
#> (Intercept)  0.03918668 0.1991352 -0.3443703  0.4388762
#> X           -1.54308670 0.5257698 -2.5815248 -0.7613898
#> 
#> $`Other parameters`
#>                  mean  std_error  lower_95  upper_95
#> c parameter 0.3024514 0.09427923 0.1067364 0.4676210
#> d parameter 0.9328742 0.02038635 0.8929056 0.9712623
#> 
#> $`Fit measures`
#>        DIC -2*LPML     WAIC
#> 1 1047.701 1047.88 1047.701

DouglasMesquita/robit documentation built on May 28, 2022, 8:30 a.m.