Railways across the world

Purpose

This work is part of an assignment of Commerce and Circulation chair at UNIMONTES and is currently under development. Feel free to ask me any questions.

Libraries

library(tidyverse)
library(sf)
library(tmap)
library(wbstats)
library(geobr)

Feature extraction

Goods transported by railways (ton/km)

railways_goods <- wb_data("IS.RRS.GOOD.MT.K6", mrnev = 1)

railways_goods <- railways_goods %>%
  arrange(-IS.RRS.GOOD.MT.K6) %>%
  rename(GOODS_T = IS.RRS.GOOD.MT.K6)

People transported by railways (person/km)

railways_people <- wb_data("IS.RRS.PASG.KM", mrnev = 1)

railways_people <- railways_people %>%
  arrange(-IS.RRS.PASG.KM) %>%
  rename(PEOPLE_T = IS.RRS.PASG.KM)

Country population (thousands)

population <- wb_data("SP.POP.TOTL", mrnev = 1)

population <- population %>%
  rename(POPULATION = SP.POP.TOTL)

Country GDP (PPP)

gdp_ppp <- wb_data("NY.GDP.PCAP.PP.CD", mrnev = 1)

gdp_ppp <- gdp_ppp %>%
  rename(GDP_PPP = NY.GDP.PCAP.PP.CD)

Dataset construction

dataset <- railways_goods %>%
  select(iso3c, GOODS_T) %>%
  full_join(
    railways_people %>%
      select(iso3c, PEOPLE_T),
    by = "iso3c"
  ) %>%
  full_join(
    population %>%
      select(iso3c, POPULATION),
    by = "iso3c"
  ) %>%
  full_join(
    gdp_ppp %>%
      select(iso3c, GDP_PPP),
    by = "iso3c"
  )

Correlations

People transported on goods transported

Although there are some exceptions, it is noticeable that there is some positive correlaetion between the goods transported and the people transported within countries.

dataset %>%
  ggplot() +
  geom_point(aes(GOODS_T, PEOPLE_T)) +
  geom_smooth(aes(GOODS_T, PEOPLE_T), method = "lm", se = FALSE) +
  annotate("text", label = paste0("Pearson: ", format(cor(dataset$GOODS_T, dataset$PEOPLE_T, use = "complete"), digits = 2)), x = 10, y = 10000) +
  scale_x_log10() +
  scale_y_log10()

Railway demand

Goods transported on GDP

dataset %>%
  ggplot() +
  geom_point(aes(GDP_PPP, GOODS_T)) +
  geom_smooth(aes(GDP_PPP, GOODS_T), method = "lm", se = FALSE) +
  annotate("text", label = paste0("Pearson: ", format(cor(dataset$GDP_PPP, dataset$GOODS_T, use = "complete"), digits = 2)), x = 2e3, y = 1e5) +
  scale_x_log10() +
  scale_y_log10()

People transported on population

dataset %>%
  ggplot() +
  geom_point(aes(POPULATION, PEOPLE_T)) +
  geom_smooth(aes(POPULATION, PEOPLE_T), method = "lm", se = FALSE) +
  annotate("text", label = paste0("Pearson: ", format(cor(dataset$POPULATION, dataset$PEOPLE_T, use = "complete"), digits = 2)), x = 1e5, y = 1e4) +
  scale_x_log10() +
  scale_y_log10()

Frequencies

Top 5 countries on goods transported

dataset %>%
  filter(!is.na(GOODS_T)) %>%
  mutate(
    iso3c = as_factor(iso3c) %>%
      fct_reorder(-GOODS_T) %>%
      fct_lump(5, w = GOODS_T)
  ) %>%
  ggplot() +
  geom_col(aes(x = iso3c, y = GOODS_T))

Top 5 countries on people transported

dataset %>%
  filter(!is.na(PEOPLE_T)) %>%
  mutate(
    iso3c = as_factor(iso3c) %>%
      fct_reorder(-PEOPLE_T) %>%
      fct_lump(5, w = PEOPLE_T)
  ) %>%
  ggplot() +
  geom_col(aes(x = iso3c, y = PEOPLE_T))

Brazil on perspective

Goods transported

Top 5 countries and Brazil

dataset %>%
  filter(!is.na(GOODS_T)) %>%
  mutate(
    iso3c = ifelse(
      iso3c == "BRA",
      "BRA",
      iso3c %>%
        as_factor() %>%
        fct_reorder(-GOODS_T) %>%
        fct_lump(5, w = GOODS_T) %>% as.character()
    ) %>%
      as_factor() %>%
      fct_reorder(-GOODS_T)
  ) %>%
  ggplot() +
  geom_col(aes(x = iso3c, y = GOODS_T))

People transported