In OlivierBinette/MSETools: Interface to Multiple Systems Estimation Software

knitr::opts_chunk$set(
  collapse = TRUE,
  message = FALSE,
  warning = FALSE,
  dev=c("png", "pdf"),
  comment = "#>",
  fig.align = "center"
)

if (!require(pacman)) install.packages("pacman")
pacman::p_load(ggplot2, dplyr, knitr, latex2exp, tibble, tidyr)
pacman::p_load_gh("OlivierBinette/pretty")

SparseMSE sensitivity {#sensitivity_SparseMSE}

We evaluate the sensitivity of sparsemse to the p-value threshold when applied to the UK dataset.

library(MSETools)

pthresh = c(seq(0, 0.01, by=0.0001), seq(0.011, 0.1, by=0.001))
datasets = list("United Kingdom" = UK, 
                "Netherlands" = Netherlands, 
                "New Orleans" = NewOrleans, 
                "Western U.S." = WesternUS, 
                "Australia" = Australia)
params = as_tibble(expand.grid(pthresh=pthresh, dataset=datasets))
param_names = as_tibble(expand.grid(pthresh=pthresh, dataset=names(datasets), stringsAsFactors=FALSE))

models = apply(params, 1, function(x) sparsemse(x$dataset, pthresh=x$pthresh))

ests = batch.estimates(models, njobs=length(models), mc.cores=20, plan="collect", job_name="SparseMSE_sensitivity", nboot=10000)

param_names %>% 
  add_column(ests=ests) %>% 
  tidyr::unnest_wider(ests) %>% 
  filter(dataset=="United Kingdom") %>% 
  ggplot() + 
  geom_ribbon(aes(x=pthresh, ymin=N.lwr, ymax=N.upr), fill="#0072B2", alpha=0.5) +
  geom_line(aes(x=pthresh, y=N.hat)) +
  expand_limits(y=0) +
  xlab("p-value threshold") +
  ylab("Estimated population size") +
  theme_minimal()