# =====================================================================================================================
# EThekwini Electricity Usage and Access                                          Andrew Collier (andrew@exegetic.biz)
#                                            Exegetic Analytics / www.exegetic.biz / info@exegetic.biz / @exegeticdata
# =====================================================================================================================

# https://edge.durban/dataset/electricity-usage

library(readxl)
library(ggplot2)
library(janitor)
library(dplyr)
library(tidyr)

# DOWNLOAD XLSX -------------------------------------------------------------------------------------------------------

FILENAME = "electricity-use-ethekwini.xlsx"

URL = "http://edge.durban/dataset/d57748e4-c446-4e7c-914d-40f754b8d77a/resource/434026e6-7b26-4dc5-9247-13d6ea31697c/download/"
#
URL = file.path(URL, FILENAME)

download.file(URL, FILENAME)

# LOAD DATA -----------------------------------------------------------------------------------------------------------

electricity <- read_xlsx(FILENAME)

View(electricity)
#
# Those column names are going to be a problem!

electricity <- clean_names(electricity)

View(electricity)

# PREPARE DATA --------------------------------------------------------------------------------------------------------

# The year column is not a number!
#
electricity <- electricity %>%
  separate(year, c("year", NA)) %>%
  mutate(year = as.integer(year))

# PLOTS ---------------------------------------------------------------------------------------------------------------

ggplot(electricity, aes(x = year, y = number_of_customers / 1000)) +
  geom_col(fill = "#3498db") +
  scale_x_continuous(NULL, breaks = seq(1980, 2020, 5)) +
  scale_y_continuous("Customers [thousands]")

ggsave("electricity-usage-customers.png", width = 6, height = 3, dpi = 150)

ggplot(electricity, aes(x = year, y = energy_k_wh_sold / 1e9)) +
  geom_col(fill = "#ffa330") +
  geom_vline(xintercept = 2007, lty = "dashed") +
  scale_x_continuous(NULL, breaks = seq(1980, 2020, 5)) +
  scale_y_continuous("Energy [tWh]")

ggsave("electricity-usage-energy.png", width = 6, height = 3, dpi = 150)

ggplot(electricity, aes(x = year, y = energy_k_wh_sold / number_of_customers / (365.25 * 24))) +
  geom_line() +
  geom_point() +
  geom_vline(xintercept = 2007, lty = "dashed") +
  scale_x_continuous(NULL, breaks = seq(1980, 2020, 5)) +
  scale_y_continuous("Average Power per Customer [kW]")

ggsave("electricity-usage-power-per-customer.png", width = 6, height = 3, dpi = 150)