In vincenzocoia/distplyr: Manipulation of Univariate Distributions

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(distplyr)
library(dplyr)
library(ggplot2)
library(tidyr)

base <- dst_empirical(-2:2)
dri <- base %>%
    graft_right(dst_norm(0, 1), breakpoint = 0, include = TRUE)
dre <- base %>%
    graft_right(dst_norm(0, 1), breakpoint = 0, include = FALSE)
dli <- base %>%
    graft_left(dst_norm(0, 1), breakpoint = 0, include = TRUE)
dle <- base %>%
    graft_left(dst_norm(0, 1), breakpoint = 0, include = FALSE)
xri <- eval_quantile(dri, at = runif(10000))
xre <- eval_quantile(dre, at = runif(10000))
xli <- eval_quantile(dli, at = runif(10000))
xle <- eval_quantile(dle, at = runif(10000))
eri <- dst_empirical(xri)
ere <- dst_empirical(xre)
eli <- dst_empirical(xli)
ele <- dst_empirical(xle)
enframe_cdf(dri, eri, at = seq(-3, 3, length.out = 1000)) %>%
    pivot_longer(cols = !.arg, names_to = "distribution", values_to = "cdf") %>%
    ggplot(aes(.arg, cdf)) +
    geom_line(aes(group = distribution, colour = distribution), alpha = 0.5) +
    theme_minimal()
enframe_cdf(dre, ere, at = seq(-3, 3, length.out = 1000)) %>%
    pivot_longer(cols = !.arg, names_to = "distribution", values_to = "cdf") %>%
    ggplot(aes(.arg, cdf)) +
    geom_line(aes(group = distribution, colour = distribution), alpha = 0.5) +
    theme_minimal()
enframe_cdf(dli, eli, at = seq(-3, 3, length.out = 1000)) %>%
    pivot_longer(cols = !.arg, names_to = "distribution", values_to = "cdf") %>%
    ggplot(aes(.arg, cdf)) +
    geom_line(aes(group = distribution, colour = distribution), alpha = 0.5) +
    theme_minimal()
enframe_cdf(dle, ele, at = seq(-3, 3, length.out = 1000)) %>%
    pivot_longer(cols = !.arg, names_to = "distribution", values_to = "cdf") %>%
    ggplot(aes(.arg, cdf)) +
    geom_line(aes(group = distribution, colour = distribution), alpha = 0.5) +
    theme_minimal()