Analyzing Water Access Points.Rmd

---
title: "Tidy Tuesday - Analyzing water access points in R"
author: "Andrew Tran"
date: "2024-03-18"
output: html_document
---

```{r setup, include=FALSE}

# Dataset from Tidy Tuesday project: https://github.com/rfordatascience/tidytuesday/blob/master/data/2021/2021-05-04/readme.md

# Install Packages
library(tidyverse)
library(scales)
library(tidytuesdayR)
library(countrycode)

theme_set(theme_light())

knitr::opts_chunk$set(echo = TRUE)
```

```{r data import}

tt <- tt_load("2021-05-04")

```

```{r data cleaning, echo=FALSE}

library(lubridate)

water <- tt$water %>%
            filter(!country_name %in% c("Peru", "Dominican Republic", "Timor-Leste"),
                   !is.na(country_name)) %>% # filtered out data without a country or in peru
            filter(between(lat_deg, -35, 37),
                   between(lon_deg, -40, 60)) %>%
            mutate(report_date = mdy(report_date),
                   install_year = ifelse(install_year > 2021, NA_real_, install_year)) %>%
            rename(lat = lat_deg,
                   lon = lon_deg,
                   country = country_name) %>%
            separate(water_tech, c("water_tech", "brand"), sep = " - ", fill = "right")

```

```{r data exploration, echo=FALSE}

# count the number of values for the status id
water %>%
  count(status_id)
# most common status id is yes

# count the number of value for each water source
water %>%
  count(water_source, sort = T)
# most common water source is Borehole

# count the nuymber of values for each water_tech value
# water tech = system used to transport water
water %>%
  count(water_tech, sort = T)

water %>%
  count(water_source, water_tech, sort = T)

# visualizing the distribution of install years
water %>%
  filter(install_year > 1980) %>% # look at recent install years
  count(install_year) %>%
  ggplot(aes(install_year, n)) +
  geom_col()

water %>%
  count(country, sort = T)

water %>%
  count(installer, sort = T)

# Pay is a free text column
water %>%
  count(pay, sort = T)

water %>%
  count(status_id, status_id, sort =  T)

```

```{r mapping}

library(ggthemes)

countries <- unique(water$country)

africa_map_data <- map_data("world") %>%
                      as_tibble() %>%
                      mutate(continent = countrycode(region, "country.name", "continent")) %>%
                      filter(continent == "Africa")

water %>%
  group_by(country) %>%
  summarise(lat = mean(lat),
            lon = mean(lon)) %>%
  ggplot(aes(lon, lat)) +
  geom_point() +
  geom_text(aes(label = country), vjust = 1, hjust = 1)

water %>%
  count(country, sort = TRUE)

water %>%
  sample_n(50000) %>% #random sample of 50000
  ggplot(aes(lon, lat)) +
  geom_polygon(aes(long, lat, group = group), 
               color = "gray", 
               fill = "white", 
               data = africa_map_data,
               size = .25) + 
  geom_point(size = .1, alpha = .25) +
  theme_map()

```

```{r Uganda}

# Wehre are recorded water wells in Uganda
# Focusing on Uganda

water_uganda <- water %>%
                  filter(country == "Uganda",
                         between(lat, -2, 4),
                         between(lon, 29, 40)) 

water_uganda %>%
      ggplot(aes(lon, lat, color = status_id)) +
      borders("world", regions = "Uganda")+ 
      geom_point(size = .1, alpha = .25) +
      theme_map() +
      scale_color_discrete(guide = guide_legend(override.aes = list(size = 2, alpha = 1)))
```

```{r}
bbox <- c(left = 29.2, bottom = -2, right = 35, top = 4.2)

uganda_map <- get_stadiamap(bbox, zoom = 9)

ggmap(uganda_map) +
  geom_point(aes(lon, lat),
             data = water_uganda, size = .1, alpha = .25)

water_uganda %>%
  count(water_source, sort = T)

water_uganda %>%
  count(pay, sort = T)


```