Script permite modelar desigualdad "natural"

ineq inevitable.R

library(tidyverse)
library(ineq)

options(scipen = 999)

theme_set(theme_classic())

base <- rep(1000, 1000)

lista <- list()

for (i in 1:100) {
  lista[[i]] <- base
  base <- base * (1 + runif(100, 0.0000001, 1))
}

data <- do.call("rbind", lista) %>% t() %>% as.data.frame()

rm(i)
# nrow(data)
# ncol(data)

data2 <- data %>% 
  gather("simulation", "value", 1:100)

data2$simulation <- str_replace_all(data2$simulation, "V", "") %>%
  as.numeric()

ginis <- data2 %>%
  group_by(simulation) %>%
  summarize(gini = ineq(value, type = "Gini"))

ggplot(ginis, aes(simulation, gini)) +
  geom_line() +
  labs(x="Nº simulación", y ="Gini")

lorenz <- data2 %>%
  group_by(simulation, value) %>%
  summarize(n = n()) %>%
  mutate(p = n/sum(n)) %>%
  mutate(acumulado.pop = cumsum(p)) %>%
  mutate(valrelat = p * value) %>%
  mutate(acumulado.val = cumsum(valrelat/sum(valrelat))) %>%
  as.data.frame()


lorenz <- lorenz[-1, ] 


ggplot(lorenz, aes(acumulado.pop, acumulado.val, color = simulation)) +
  geom_line() +
  geom_abline(color = "red") +
  scale_color_continuous(trans= "reverse",
                         breaks = seq(0, 100, 25)) +
  labs(x="Población relativa acumulada", 
       y="Recurso relativo acumulado",
       color ="Simulación")