---
title: "Session 5 — Outliers"
author: "Prof. Dae-Jin Lee (daelee@faculty.ie.edu)"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, message = FALSE, warning = FALSE)
library(tidyverse)
ci <- read_csv("data/country_indicators.csv")
```

```{r}
r_all <- cor(ci$gdp_per_capita, ci$life_expectancy)

ci |>
  mutate(r_without = map_dbl(country, function(ctry) {
    d <- filter(ci, country != ctry)
    cor(d$gdp_per_capita, d$life_expectancy)
  }),
  delta = r_without - r_all) |>
  select(country, gdp_per_capita, life_expectancy, r_without, delta) |>
  arrange(delta) |>
  head(5)
```

**Interpret:** Countries with the most negative `delta` pull $r$ upward when included.