Posner_graphics.Rmd

---
title: "Pardee Trends: Graphics"
author: "Mickey Rafa"
date: "July 2017"
output:
  html_notebook:
    toc: yes
    fig.width: 7
    fig.align: "center"
  html_document:
    toc: yes
    toc_depth: 3
    fig.height: 6.2
    fig.align: "center"
---

```{r setup, include=FALSE}
library(tidyverse)
library(stringr)
library(gridExtra)
library(scales)
library(extrafont)
library(cowplot)
library(tweenr)
library(gganimate)
library(grid)
library(magick)
```

```{r pardee.theme, include=TRUE}
pardee.theme <- theme(
  plot.title = element_text(family="Gill Sans MT", color="#871B31", face="bold", size=36, hjust=0), 
  plot.subtitle = element_text(family="Gill Sans MT", color="black", size=14, hjust=0), 
  axis.title = element_text(family="Gill Sans MT", color="black", face="bold", size=12), 
  axis.text = element_text(family="Gill Sans MT", color="black", size=22),
  axis.title.x = element_text(family="Gill Sans MT", color="black", size=22),
  axis.title.y = element_text(family="Gill Sans MT", color="black", size=22),
  plot.caption = element_text(family="Gill Sans MT", color="black", size=12), 
  panel.border = element_blank(),
  panel.background =  element_blank(),
  panel.grid.minor = element_blank(),
  panel.grid.major = element_blank(),
  legend.title=element_blank(), 
  legend.text = element_text(family="Gill Sans MT", color="black", size=20, hjust = 0),
  legend.spacing.x = unit(.3, 'cm'), 
  legend.position = 'top',
  strip.text.x = element_text(margin = margin(.3,0,.3,0, "cm")), 
  strip.text = element_text(family="Gill Sans MT", color="black", size=16, face="bold"))

line_2017 <- geom_vline(xintercept = 2017, linetype = 'longdash')

pt_groups <- read.csv('pardee_trends_groups.csv')
pardee_group <- pt_groups %>% filter(main_group == 'Pardee Group')

p_trends_cols <- c('POPGT65' = '#871B31', 'POP0TO5' = "#D4B75A",
                   'Sub-Saharan Africa' = 'steelblue3', 'Rest of World' = 'darkgray')
```

```{r logo, include=TRUE}
logo <- image_read(".//images/pardee_logo.png")
logo_ext <- image_read(".//images/pardee_logo_extended.png")
```


```{r read_data, include=TRUE}
df <- read_csv('.//data/ifs_results_posner.csv') %>% 
  gather(year, val, 5:145) %>% 
  mutate(year = as.numeric(year), 
         variable = ifelse(str_detect(variable, "\\["), str_sub(variable, end=-4), variable))
```

```{r fig.height=6.2}
p <- df %>% 
  filter(variable %in% c('POPGT65', 'POP0TO5'),
         year <= 2060) %>% 
  group_by(variable, year) %>% 
  summarise(val = sum(val, na.rm = TRUE)) %>% 
  ggplot(.) + 
  geom_line(aes(x=year,
                y=val / 1000,
                group=variable,
                color=variable), size = 2) + 
  ggtitle('World population is aging fast') + 
  line_2017 + 
  labs(y='',
       x='') + 
  annotate(geom = "text", x=2007, y=.16,
           label = expression('' %<-%'Estimate'), color="black", size=6, family="Gill Sans MT") + 
  annotate(geom = "text", x=2028, y=.15,
           label = expression('Projection' %->% ''), color="black", size=6, family="Gill Sans MT") + 
  scale_x_continuous(breaks = c(seq(1960, 2060, by=20))) + 
  scale_y_continuous(position = "right") + 
  scale_color_manual(values = p_trends_cols,
                     labels = c('POP0TO5' = 'Under 5',
                                'POPGT65' = '65 and over')) + 
  pardee.theme + 
  theme(legend.position = c(.05, .87),
        legend.key.height = unit(.5, "inches"))
```

```{r fig.height=6.2}
p
grid::grid.raster(logo, x=.88, y=.93, width = unit(1, 'inches'))
```


```{r}
ggsave(".//graphics/testfig_1.pdf", plot= p,
       height = 6.2, width = 10, family="Gill Sans MT", device=cairo_pdf, dpi = 300)
```

```{r fig.height=6.2}
p + 
  labs(caption = '\n')
grid::grid.raster(logo_ext, x=.8, y=.07, width = unit(2.5, 'inches'))
```



```{r fig.height=6.2}
df %>% 
  filter(year %in% c(2019, 2060),
         variable == 'POP0TO5') %>% 
  spread(year, val) %>% 
  mutate(diff = `2060` - `2019`) %>% 
  arrange(-diff) %>% 
  top_n(8, diff) %>% 
  ggplot(.) + 
  geom_bar(aes(x=reorder(country, diff),
               y=diff), stat='identity', color='darkgray', fill='seagreen') +
  ggtitle('Change in under 5 populations by country from 2019 to 2060', 
          subtitle = 'Top 8') + 
  labs(x='',
       y='Growth in Under 5 population (millions)') + 
  coord_flip() + 
  scale_x_discrete(labels = function(x) str_wrap(x, width = 15)) + 
  pardee.theme + 
  scale_fill_manual(values = p_trends_cols) + 
  theme(plot.title = element_text(size=19),
        axis.text = element_text(size=16),
        axis.title.x = element_text(size=16))
```

```{r fig.height=7}
df_health <- df %>% 
  filter(variable %in% c('Commun', 'NonCommun', 'Injuries'),
         #country %in% c('WB High-Income Economies', 'WB Low-Income Economies'), 
         year <= 2060, 
         !is.na(val))

df_health$country <- factor(df_health$country, levels = c('WB Low-Income Economies', 'WB LowMidIncome Economies',
                                                          'WB UpMidIncome Economies', 'WB High-Income Economies'))

df_health %>% 
  ggplot(.) + 
  geom_line(aes(x=year,
                y=val,
                color=variable), size = 2) + 
  labs(y='Deaths per thousand\n',
       x='') + 
  facet_wrap(~country,
             labeller = as_labeller(c('WB Low-Income Economies' = 'Low Income',
                                      'WB LowMidIncome Economies' = 'Low-Middle Income',
                                      'WB UpMidIncome Economies' = 'Upper-Middle Income',
                                      'WB High-Income Economies' = 'High Income'))) + 
  scale_color_manual(values = c("red2", "green3", "blue"),
                     labels = c('Communicable', 'Injuries & Accidents', 'Non-communicable')) + 
  pardee.theme + 
  theme(axis.text = element_text(size=16))
```

```{r fig.height=6.2}
demog_trans <- df %>% 
  left_join(x=.,
            y=pardee_group,
            by=c('country' = 'Country')) %>% 
  filter(variable %in% c('CBR', 'CDR')) %>% 
  filter(country %in% c('Japan', 'Uganda', 'Brazil')) %>% 
  filter(year <= 2060 ) %>% 
  spread(variable, val)
```


```{r fig.width=10}
demog_trans_edit <- demog_trans %>%
  arrange(country, year) %>%
  select(CDR, CBR, year, country, sub_group) %>%
  rename(x=CDR,y=CBR,time=year,id=country) %>%
  mutate(ease="linear")

demog_trans_tween <- tween_elements(demog_trans_edit,
                                  "time", "id", "ease", nframes = 100) %>%
  mutate(year = round(time), country = .group) %>%
  left_join(demog_trans, by=c("country","year","sub_group"))

animate_1 <- demog_trans_tween %>% 
  ggplot(.,
         aes(x=x, y=y, frame = year)) +
  geom_point(aes(color=sub_group), size = 3) +
  ggtitle('Demographic Transition') + 
  labs(x='Crude death rate\n',
       y='Crude birth rate\n') + 
  pardee.theme
  #scale_alpha_discrete(range = c('China' = 1,
   #                               'Other' = .25)) + 
  #theme(legend.position = 'top') + 
  #guides(size='none', alpha='none')

gganimate(animate_1, interval=0.1, ani.width = 1000, ani.height = 700, ".//graphics/test.gif")
```

```{r}
(demog_trans_tween$country <- factor(demog_trans_tween$country, levels = c('Uganda',
                                                                          'Brazil',
                                                                          'Japan')))


(animate_2 <- demog_trans_tween %>% 
  ggplot(.,
         aes(x=x,
             y=y,
             color=country)) +  
  geom_path(aes(group=country),
            size=1.75,
            alpha=.3) +
  geom_point(aes(frame = year), size=7) + 
  #annotate("text", x=16, y=10, label="Japan", family="Gill Sans MT", size=15) + 
  ggtitle('Demographic Transition:',
          subtitle = '') + 
  labs(x='Crude death rate\n',
       y='Crude birth rate\n') + 
  scale_color_manual(values = c("steelblue", "tomato3", "green3"), 
                     labels = c("Uganda", "Brazil", "?")) + 
  #guides(color=guide_legend()) + 
  pardee.theme + 
   theme(plot.title = element_text(size=50),
         legend.text = element_text(size=35),
         axis.text.x = element_text(size=30),
         axis.text.y = element_text(size=30),
         axis.title = element_text(25),
         legend.spacing.x = unit(1.2, 'cm')))
```

```{r}
gganimate(animate_2, interval=0.2, ani.width = 1000, ani.height = 700, ".//graphics/test_5.gif")
```

```{r fig.height=6.2}
(p_2 <- df %>% 
  filter(variable == 'INCOMELT190LN',
         year %in% c(1990, 2015, 2040)) %>% 
  ggplot(.) + 
  geom_bar(aes(x=as.factor(year),
               y=val / 1000,
               fill=country), 
           stat='identity', position='stack', color='darkgray') + 
  #ggtitle('Extreme poverty headcount: Sub-Saharan Africa vs. Rest of World') + 
  labs(x='Year',
       y='Population below $3.10 / day (billions)\n') + 
  scale_fill_manual(values = p_trends_cols) + 
  pardee.theme + 
  theme(axis.title.y = element_text(size=16)))
```




```{r fig.height=7, fig.width=10}
ggplot() + 
  ggtitle('Demographic Transition\n') + 
  labs(x='Crude death rate\n(Deaths per thousand population)\n',
       y='Crude birth rate\n(Births per thousand population)\n') + 
  pardee.theme + 
  theme(plot.title = element_text(size=30))

ggsave(".//graphics/test_start.png", width = 10, height = 7)
```