---
title: 'Analyzing inequality in access to opportunities'
date: "`r Sys.Date()`"
output: rmarkdown::html_vignette
urlcolor: blue
vignette: >
  %\VignetteIndexEntry{Analyzing inequality in access to opportunities} 
  %\VignetteEngine{knitr::rmarkdown} 
  \usepackage[utf8]{inputenc}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  eval = identical(tolower(Sys.getenv("NOT_CRAN")), "true"),
  out.width = "100%"
)
```


Social and racial inequalities in access to opportunities

```{r, message = FALSE, eval = TRUE, warning=FALSE}
# load libraries
library(aopdata)
library(sf)
library(ggplot2)
library(data.table)
library(units)
```


## Download accessibility data

```{r, message = FALSE, eval = TRUE}
df <- aopdata::read_access(
  city = 'Curitiba',
  mode = 'public_transport',
  peak = TRUE,
  year = 2019,
  showProgress = FALSE
  )
```


## Inequality in access to job opportunities by income decile

```{r, message = FALSE, eval = !is.null(df)}
ggplot() +
  geom_boxplot(data=subset(df, !is.na(R003)),
               aes(x = factor(R003), y=CMATT60/1000, color=factor(R003))) +
  scale_color_brewer(palette = 'RdBu') +
  labs(title='Distribution of the number of jobs accessible', color="Income\ndecile",
       subtitle='by public transport in less than 60 min. by income decile',
       x='Income decile', y="N. of jobs accessible\n(thousands)") +
  scale_x_discrete(labels=c("D1 Poorest", paste0('D', 2:9), "D10 Wealthiest")) +
  theme_minimal()
```


## Palma ratio

```{r, message = FALSE, eval = !is.null(df)}
# average access of the wealthiest 10%
avg_access_10p_wealthiest <- df[ R003==10, weighted.mean(x=CMATT60, w=P001, na.rm=T)]

# average access of the poorest 40%
avg_access_40p_poorest <- df[ R003<=4, weighted.mean(x=CMATT60, w=P001, na.rm=T)]

# Palma ratio
palma_ratio <- avg_access_10p_wealthiest / avg_access_40p_poorest                
palma_ratio 
```

This means that the 10% wealthiest population could access by public transport 
on average 2.6 times more job opportunites than the 40% 
poorest people in less than 60 min.



## Inequality in travel time to closes hospital

```{r, message = FALSE, eval = !is.null(df)}
# replace Inf travel time with 120
df[, TMISA := fifelse(TMISA==Inf, 120, TMISA)]

# calculate average travel time by race
df[, .(average = weighted.mean(x=TMISA, w=P001, na.rm=T),
       white   = weighted.mean(x=TMISA, w=P002, na.rm=T),
       black   = weighted.mean(x=TMISA, w=P003, na.rm=T))]

# calculate average travel time by income
temp <- df[, .(average = weighted.mean(x=TMISA, w=P001, na.rm=T)), by=R003]
temp <- na.omit(temp)

ggplot() + 
  geom_point(data=temp, aes(y=average, x=factor(R003))) +
  labs(x='Income decile', y='Avg. travel time to\nclosest hospital') +
  theme_minimal()
```