2021_01_05_transit_costs.Rmd

---
title: "TidyTemplate"
date: 2021-01-05
output: html_output
---

# TidyTuesday

Join the R4DS Online Learning Community in the weekly #TidyTuesday event!
Every week we post a raw dataset, a chart or article related to that dataset, and ask you to explore the data.
While the dataset will be “tamed”, it will not always be tidy! As such you might need to apply various R for Data Science techniques to wrangle the data into a true tidy format.
The goal of TidyTuesday is to apply your R skills, get feedback, explore other’s work, and connect with the greater #RStats community!
As such we encourage everyone of all skills to participate!

```{r setup, include=FALSE}

knitr::opts_chunk$set(echo = TRUE)

library(tidyverse)
library(tidytuesdayR)
library(scales)
library(glue)
theme_set(theme_light())

```

# Load the weekly Data

Dowload the weekly data and make available in the `tt` object.

```{r Load}

tt <- tt_load("2021-01-05")

library(countrycode)

transit_cost <- tt$transit_cost %>%
  filter(!is.na(e)) %>%
  mutate_at(vars(start_year, end_year, real_cost), as.numeric) %>%
  mutate(country_code = ifelse(country == "UK", "GB", country),
         country = countrycode(country_code, "iso2c", "country.name"),
         tunnel_per = tunnel / length,
         rr = ifelse(rr, "Railroad", "Not Railroad"))
```

```{r}
transit_cost %>%
  count(city, country, sort = TRUE)

transit_cost %>%
  filter(country == "United States") %>%
  mutate(line = fct_reorder(line, year)) %>%
  ggplot(aes(xmin = start_year, xmax = end_year, y = line,
             color = city,
             size = real_cost)) +
  geom_errorbarh(height = .1) +
  labs(x = "Year",
       y = "",
       color = "City")

transit_cost %>%
  ggplot(aes(cost_km_millions)) +
  geom_histogram() +
  scale_x_continuous(labels = dollar) +
  labs(x = "Cost / KM (Millions of USD)")

transit_cost %>%
  filter(!is.na(cost_km_millions),
         tunnel_per == 1) %>%
  mutate(country = fct_lump(country, 10)) %>%
  add_count(country) %>%
  mutate(country = glue("{ country } ({ n })"),
         country = fct_reorder(country, cost_km_millions)) %>%
  ggplot(aes(cost_km_millions, country)) +
  geom_boxplot() +
  scale_x_continuous(labels = dollar) +
  labs(x = "Cost / KM (Millions of USD)",
       y = "")
```

```{r}
transit_cost %>%
  filter(!is.na(cost_km_millions),
         tunnel_per == 1,
         country == "China") %>%
  mutate(city = fct_lump(city, 10)) %>%
  add_count(city) %>%
  mutate(city = glue("{ city } ({ n })"),
         city = fct_reorder(city, cost_km_millions)) %>%
  ggplot(aes(cost_km_millions, city)) +
  geom_boxplot() +
  scale_x_continuous(labels = dollar) +
  labs(x = "Cost / KM (Millions of USD)",
       y = "") +
  expand_limits(x = 0)

transit_cost %>%
  filter(!is.na(cost_km_millions),
         tunnel_per == 1,
         country == "China")
```


```{r}
transit_cost %>%
  filter(country == "China",
         city == "Shanghai",
         !is.na(start_year),
         !is.na(end_year)) %>%
  mutate(city = fct_lump(city, 5)) %>%
  mutate(line = fct_reorder(line, year)) %>%
  ggplot(aes(xmin = start_year, xmax = end_year, y = line,
             color = city,
             size = real_cost)) +
  geom_errorbarh(height = .1) +
  labs(x = "Year",
       y = "",
       color = "City")

transit_cost %>%
  filter(tunnel_per == 1,
         end_year <= 2020,
         country == "China") %>%
  group_by(year = (year %/% 5) * 5) %>%
  summarize(median_cost_km = median(cost_km_millions),
            n = n()) %>%
  ggplot(aes(year, median_cost_km)) +
  geom_line() +
  geom_point(aes(size = n))

transit_cost %>%
  filter(tunnel_per == 1,
         end_year <= 2020,
         country == "China") %>%
  mutate(year = (year %/% 5) * 5,
         city = fct_lump(city, 5)) %>%
  ggplot(aes(year, cost_km_millions, group = year)) +
  geom_boxplot(outlier.size = -1) +
  geom_jitter(aes(color = city), height = 0, width = 1) +
  expand_limits(y = 0) +
  labs(y = "Cost / km (Real USD, Millions)",
       x = "Year",
       title = "Cost distribution / km in China")

transit_cost %>%
  filter(tunnel_per == 1,
         end_year <= 2020,
         country == "China") %>%
  mutate(city = fct_lump(city, 4)) %>%
  ggplot(aes(stations / length, cost_km_millions, size = length,
             color = city)) +
  geom_point() +
  expand_limits(x = 0, y = 0) +
  labs(x = "Stations / km", "Cost / kilometer",
       y = "Cost / km")
```

Conclusions: Prices in China ranged from slightly above 200M/KM to a bit under 200M/KM from 2000 to 2015, and have since risen a bit.

cost_km_millions comes from real_cost / length (NOT completed length)

No strong relationship between stations/km and cost/km

```{r}
transit_cost %>%
  ggplot(aes(cost_km_millions, real_cost / length)) +
  geom_point()
```