recreates.Rmd

---
title: "Recreating figures from Storytelling with Data"
author: "Ashleigh Wilson"
date: '2022-02-24'
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

library(tidyverse)
library(readxl)
library(janitor)
library(scales)
library(lemon)
library(ggtext)
library(grid)
library(gridExtra)
library(gridtext)

source("themes/my_swd_theme.R")
```

Reading in excel spreadsheet. Use the range argument to specify the exact cells to import.

```{r}
# meh, don't want to do this graph again.

# data_2_8_og <- read_excel("data/2.8_EXERCISE.xlsx", range = "B5:F15")
# data_2_8 <- clean_names(data_2_8_og)

# data_2_8[10,1] <- "Dec-19" # just changing value name
# data_2_8 <- data_2_8 %>%
#   rename(year = x1) %>%
#   mutate(year = as_factor(year))
```

------------------------------------------------------------------------

## Chapter 2 Figures

### 2.7

```{r}
data_0207_OG <- read_csv("data/fig_0207_data.csv")
```

```{r}
data_0207 <- clean_names(data_0207_OG)
```

Graph:

```{r}
data_0207 <- data_0207 %>%
  mutate(
    cost_per_mile = parse_number(cost_per_mile), # need to convert chars with dollar sign $ by using parse_number()
    avg_miles = mean(miles_driven),
    avg_cost = mean(cost_per_mile),
    color = if_else(cost_per_mile < avg_cost, SILVER, DARKORANGE)
  )

avg_miles2 <- mean(data_0207$miles_driven)
avg_cost2 <- mean(data_0207$cost_per_mile)

ggplot(data_0207) +
  geom_point(aes(miles_driven, cost_per_mile, color = color), size = 2) +
  scale_color_identity() + # can use when data has column values that represents aesthetics (which it does with the color column values)
  scale_y_continuous(breaks = seq(0, 3, .5), limits = c(0, 3), labels = scales::dollar_format()) +
  scale_x_continuous(limits = c(0, 4000), labels = scales::label_comma()) + # original was 1000 to 3000. Changing it also shrinks the plot area
  geom_hline(yintercept = avg_cost2, linetype = "longdash") +
  geom_point(aes(avg_miles2, avg_cost2), size = 4) +
  geom_label(aes(avg_miles2, avg_cost2), label = "AVG", hjust = 1.25, size = 4.5, label.size = 0) +
  labs(
    title = "Cost per mile by miles driven",
    y = "Cost per mile",
    x = "Miles driven per month"
  ) +
  my_swd_theme() +
  coord_capped_cart(bottom = "right", left = "top") # this function won't work unless the theme I use has panel.border = element_blank() and axis.line = element_line()
```

### 2.11

```{r}
data_0211_OG <- read_csv("data/fig_0211_data.csv")
```

```{r}
data_0211_1 <- data_0211_OG %>%
  rename(survey_category = ...1) %>%
  pivot_longer(cols = c("2014", "2015"), names_to = "year", values_to = "percent") %>%
  mutate(
    percent = parse_number(percent),
    # Relevel the categories to ensure that "Career development" is drawn last and displays on top of the other categories
    survey_category = fct_relevel(factor(survey_category), "Peers", "Culture", "Work environment", "Leadership", "Rewards & recognition", "Perf management", "Career development"),
    color = if_else(survey_category == "Career development", TOMATO, DIMGRAY)
  )

# below code won't work to make the graph
data_0211_2 <- data_0211_OG %>%
  rename(survey_category = ...1) %>%
  mutate(
    `2014` = parse_number(`2014`),
    `2015` = parse_number(`2015`)
  )
```

```{r}
ggplot(data_0211_1) +
  # must make survey_category a factor before I can use as group aes
  geom_line(aes(year, percent, group = survey_category, color = color), size = 1) +
  geom_point(aes(year, percent, color = color), size = 3) +
  scale_color_identity() + # always need this when I have color column with hex codes
  scale_x_discrete(expand = expansion(1.9, 0), limits = c("2014", "2015"), labels = c(2014, 2015)) +
  geom_text(aes(year, percent, label = paste0(percent, "%"), color = color), nudge_x = -.2, data = data_0211_1 %>% filter(year == 2014) %>% filter(survey_category != "Career development")) +
  geom_text(aes(year, percent, label = paste0(percent, "%"), color = color), nudge_x = .2, data = data_0211_1 %>% filter(year == 2015) %>% filter(survey_category != "Career development")) +
  geom_text(aes(year, percent, label = survey_category, color = color), nudge_x = -.4, data = data_0211_1 %>% filter(year == 2014) %>% filter(survey_category != "Career development"), hjust = 1) +
  geom_text(aes(year, percent, label = paste0(percent, "%"), fontface = "bold", color = color), nudge_x = -.2, data = data_0211_1 %>% filter(year == 2014) %>% filter(survey_category == "Career development")) +
  geom_text(aes(year, percent, label = paste0(percent, "%"), fontface = "bold", color = color), nudge_x = .2, data = data_0211_1 %>% filter(year == 2015) %>% filter(survey_category == "Career development")) +
  geom_text(aes(year, percent, label = survey_category, color = color), fontface = "bold", nudge_x = -.4, data = data_0211_1 %>% filter(year == 2014) %>% filter(survey_category == "Career development"), hjust = 1) +
  my_swd_theme() +
  theme(
    axis.line.y = element_blank(),
    axis.line.x = element_line(size = .1),
    axis.text.y = element_blank(),
    axis.title.x = element_text(hjust = .5),
    axis.title.y = element_blank(),
    axis.ticks.y = element_blank(),
    axis.ticks.x = element_line(color = LIGHTGRAY),
    plot.title = element_text(hjust = .1, vjust = -4),
    plot.subtitle = element_text(hjust = .208, vjust = -10, size = 10, color = DIMGRAY),
  ) +
  labs(
    title = "Employee feedback over time",
    subtitle = "Survey category | Percent favorable",
    x = "survey year"
  ) +
  coord_capped_cart(bottom = "both") # makes the x axis line only connect from 2014 - 2015
```

### 2.19

```{r}
data_0219_OG <- read_csv("data/fig_0219_data.csv")
```

```{r}
data_0219 <- data_0219_OG %>%
  rename(survey_item = ...1) %>%
  clean_names() %>%
  pivot_longer(cols = c("strongly_disagree":"strongly_agree"), names_to = "rating", values_to = "percent") %>%
  mutate(
    total = parse_number(total),
    percent = parse_number(percent),
    rating = fct_relevel(factor(rating), "strongly_disagree", "disagree", "neutral", "agree", "strongly_agree"),
    rating = fct_rev(rating)
  )

color_scale <- c(
  "strongly_disagree" = DARKGRAY,
  "disagree" = DARKGRAY,
  "neutral" = LIGHTGRAY,
  "agree" = BLUEGRAY,
  "strongly_agree" = BLUEGRAY
)
```

```{r}
ggplot(data_0219, aes(x = percent, y = fct_rev(survey_item), fill = rating)) +
  geom_col(color = "white", width = 0.8) +
  scale_fill_manual(values = color_scale, guide = "none") +
  scale_x_continuous(position = "top", labels = scales::label_percent(scale = 1)) +
  my_swd_theme() +
  theme(
    axis.line.y = element_blank(),
    axis.line.x = element_line(color = DIMGRAY, size = .2),
    axis.title.y = element_blank(),
    axis.title.x = element_text(color = DARKGRAY, size = 10, hjust = .035),
    axis.text.y = element_text(hjust = 6, size = 10),
    axis.ticks.y = element_blank(),
    axis.ticks.x = element_line(size = .2, color = DIMGRAY),
    plot.margin = margin(.2, .2, .2, .2, "cm"),
    plot.title = element_text(hjust = .04, margin = margin(t = 0, r = 0, b = 10, l = 0, unit = "pt")),
    plot.subtitle = element_markdown(hjust = .41, size = 10) # element_markdown is from the ggtext package. Allows me to set markdown text in plot (See subtitle).
  ) +
  labs(
    title = "Survey results",
    subtitle = paste0(
      "<span style='color:", color_scale[1], "'>", "**Strongly Disagree**", "</span>", " | ",
      "<span style='color:", color_scale[2], "'>", "**Disagree**", "</span>", " | ",
      "<span style='color:", color_scale[3], "'>", "**Neutral**", "</span>", " | ",
      "<span style='color:", color_scale[4], "'>", "**Agree**", "</span>", " | ",
      "<span style='color:", color_scale[5], "'>", "**Strongly Agree**", "</span>"
    ),
    x = "Percent of total"
  ) +
  coord_capped_cart(top = "both")
```

### 4.14

```{r}
data_0414_OG <- read_csv("data/fig_0414_data.csv")
```

```{r}
data_0414 <- data_0414_OG %>%
  rename(ticket_type = ...1) %>%
  pivot_longer(cols = -ticket_type, names_to = "month", values_to = "amount") %>%
  mutate(
    ticket_type = as_factor(ticket_type),
    month = as_factor(month)
  )

# the below is a grob object from the grid_text package. This was the only way to add a markdown annotation text box to the graph.
graph_annotation <- grobTree(richtext_grob(
  "<span style='background_color:white'><b>2 employees quit in May.</b> Ticket volume dramatically <br> decreased in the following two months then increased after<br> hiring two new employees in July. Volume fluctuated in the<br> subsequent months.</span>",
  x = .58, y = .93, gp = gpar(col = GRAY, fontsize = 12), hjust = 0, box_gp = gpar(col = "white", fill = "white"), padding = margin(.4, 0, 0, 0, "in")
))
```

Graph

```{r}
ggplot(data_0414, aes(x = month, y = amount, group = ticket_type, color = ticket_type)) +
  geom_line(size = 1) +
  geom_point(size = 2.5, data = data_0414 %>% filter(month %in% c("Aug", "Sep", "Oct", "Nov", "Dec"))) + # seems I can use data arg in any geom to filter for specific data to be shown
  geom_vline(xintercept = "May", color = GRAY) + # the grob textbox will cover the top part of the vline, giving it that 'cutoff' appearance.
  # geom_linerange(x = "May", ymin = 0, ymax = 240) +
  scale_color_manual(values = c(GRAY, BLUEGRAY)) +
  scale_y_continuous(limits = c(0, 300), breaks = seq(0, 300, 50)) +
  scale_x_discrete(expand = expansion(add = c(.02, 3))) +
  geom_text(aes(label = amount), nudge_y = 18, data = data_0414 %>% filter(ticket_type != "Ticket Volume Processed") %>% filter(month %in% c("Aug", "Sep", "Oct", "Nov", "Dec"))) +
  geom_text(aes(label = amount), nudge_y = -18, data = data_0414 %>% filter(ticket_type != "Ticket Volume Received") %>% filter(month %in% c("Aug", "Sep", "Oct", "Nov", "Dec"))) +
  annotate("text", x = 13.5, y = 180, size = 5, color = GRAY, label = "bold(Received)", parse = TRUE) +
  annotate("text", x = 13.68, y = 142, size = 5, color = BLUEGRAY, label = "bold(Processed)", parse = TRUE) +
  annotation_custom(graph_annotation, xmin = -9) + # used with custom grob
  coord_capped_cart(bottom = "right", left = "top") +
  labs(
    title = "Ticket volume over time"
  ) +
  my_swd_theme() +
  theme(
    legend.position = "none",
    plot.margin = margin(t = .5, r = .5, b = .5, l = .5, unit = "cm"),
    plot.background = element_rect(color = "white"),
    axis.title.x = element_blank(),
    axis.title.y = element_blank(),
    axis.line = element_line(size = .2),
    axis.ticks.x = element_blank(),
  )
```

### 5.4

```{r}
data_0504_OG <- read_csv("data/fig_0504_data.csv")
```

```{r}
data_0504 <- data_0504_OG %>%
  pivot_longer(cols = "2008":"2012", names_to = "year", values_to = "number") %>%
  filter(Category != "All") %>%
  mutate(
    Category = as_factor(Category),
    color = if_else(Category == "Bachelor's degree or more", ORANGE, SILVER)
  )
```

Graph

```{r}
ggplot(data_0504, aes(year, number, group = Category, color = color)) +
  geom_line(size = 1.5) +
  geom_point(size = 3, data = data_0504 %>% filter(year %in% c("2008", "2012"))) +
  scale_color_identity() +
  scale_x_discrete(expand = expansion(add = c(5, 1))) +
  scale_y_continuous(limits = c(5, 65)) + # setting limits can `zoom-in or out` of the plot, so to speak. Gives me more space below the below most line.
  geom_text(aes(label = number), size = 6.5, nudge_x = .3, nudge_y = .4, data = data_0504 %>% filter(year == "2012") %>% mutate(number = round(number))) +
  geom_text(aes(label = number), size = 6.5, nudge_x = -.15, nudge_y = .4, hjust = 1, data = data_0504 %>% filter(year == "2008") %>% mutate(number = round(number))) +
  geom_text(aes(label = Category), size = 6, nudge_x = -.7, nudge_y = .2, hjust = 1, data = data_0504 %>% filter(year == "2008", Category != "Bachelor's degree or more")) + # hjust will line the Categories neatly
  geom_text(aes(label = Category), size = 6, nudge_x = -.7, nudge_y = .4, hjust = 1, data = data_0504 %>% filter(year == "2008", Category == "Bachelor's degree or more")) +
  labs(
    title = "New marriage rate by education",
    subtitle = paste0("<span>Number of <b>newly</b> married adults per 1,000 marriage eligible adults</span>"),
    caption = "Note: Marriage eligibility includes newly married plus those widowed or divorced.\nSource: US Census\nAdapted from PEW RESEARCH CENTER"
  ) +
  coord_capped_cart(bottom = "both") + # I think this needs to always be at the bottom of the ggplot call.
  my_swd_theme() +
  theme(
    axis.title.y = element_blank(),
    axis.title.x = element_blank(),
    axis.line.y = element_blank(),
    axis.text.y = element_blank(),
    axis.ticks.y = element_blank(),
    plot.subtitle = element_markdown(size = 10)
  )
```

### 5.13

```{r}
data_0513_OG <- read_csv("data/fig_0513_data.csv")
```

```{r}
data_0513 <- data_0513_OG %>%
  pivot_longer(cols = c(`US Population`, `Our Customers`), names_to = "customer_type", values_to = "number") %>%
  mutate(
    Segment = as_factor(Segment),
    customer_type = as_factor(customer_type),
    num_color = case_when(
      Segment == "Segment 3" ~ WHITE,
      Segment == "Segment 4" ~ WHITE,
      Segment == "Segment 5" ~ WHITE,
      TRUE ~ BLACK
    )
  )

data_0513_color_values <- c(
  "Segment 1" = SILVER,
  "Segment 2" = SILVER,
  "Segment 3" = BLUEGRAY,
  "Segment 4" = BLUEGRAY,
  "Segment 5" = BLUEGRAY,
  "Segment 6" = SILVER,
  "Segment 7" = SILVER
)

percent_30_grob <- grobTree(richtext_grob(
  "<b>30%</b>",
  gp = gpar(fontsize = 20, col = BLUEGRAY)
))

percent_50_grob <- grobTree(richtext_grob(
  "<b>50%</b>",
  gp = gpar(fontsize = 20, col = BLUEGRAY)
))

# reference to the below solutions (line_*_grob): https://rdrr.io/r/grid/grid.segments.html
line_30_grob <- segmentsGrob(
  x0 = unit(.43, "npc"),
  y0 = unit(.525, "npc"),
  x1 = unit(.43, "npc"),
  y1 = unit(.258, "npc"),
  default.units = "npc",
  arrow = NULL, name = NULL, gp = gpar(), vp = NULL
)

line_50_grob <- segmentsGrob(
  x0 = unit(.865, "npc"),
  y0 = unit(.67, "npc"),
  x1 = unit(.865, "npc"),
  y1 = unit(.22, "npc"),
  default.units = "npc",
  arrow = NULL, name = NULL, gp = gpar(), vp = NULL
)
```

Graph

```{r}
ggplot(data_0513, aes(customer_type, number, fill = Segment)) +
  # scale_fill_identity() +
  scale_fill_manual(values = data_0513_color_values) +
  geom_col(color = "white", width = .5) +
  scale_x_discrete(expand = expansion(add = c(.7, .6))) +
  geom_text(aes(label = Segment, color = Segment), size = 4.1, position = position_stack(vjust = .55), hjust = 2.2, data = data_0513 %>% filter(customer_type == "US Population")) +
  scale_color_manual(values = data_0513_color_values) +
  geom_text(aes(label = paste0(number * 100, "%"), parse = TRUE), color = WHITE, position = position_stack(vjust = .6)) +
  annotation_custom(percent_30_grob, xmin = .2, ymin = -.29) +
  annotation_custom(percent_50_grob, xmin = 2.2, ymin = -.2) +
  # geom_linerange(x = "US Population", ymin = 0, ymax = .52, size = 6) +
  annotation_custom(line_30_grob) +
  annotation_custom(line_50_grob) +
  labs(
    title = "Distribution by customer segment"
  ) +
  my_swd_theme() +
  theme(
    axis.title.x = element_blank(),
    axis.title.y = element_blank(),
    axis.line.y = element_blank(),
    axis.ticks.y = element_blank(),
    axis.text.y = element_blank()
  )
```

### 6.1

```{r}
data_0601_OG <- read_csv("data/fig_0601_data.csv")
```

```{r}
data_0601 <- data_0601_OG %>%
  clean_names() %>%
  mutate(
    last_year = parse_number(last_year),
    progress_to_date = parse_number(progress_to_date)
  ) %>%
  pivot_longer(cols = c(last_year, progress_to_date), names_to = "money_raised_type", values_to = "amount")

color_scale_0601 <- c(
  "last_year" = LIGHTSTEELBLUE,
  "progress_to_date" = BLUEGRAY
)

progress_grob_0601 <- grobTree(richtext_grob(
  "<b>$33,967</b>",
  gp = gpar(col = BLUEGRAY, fontsize = 12)
))

last_year_grob_0601 <- grobTree(richtext_grob(
  "<b>$51,400</b>",
  gp = gpar(col = LIGHTSTEELBLUE, fontsize = 8)
))
```

Graph

```{r}
ggplot(data_0601, aes(days_since_launch, amount, color = money_raised_type, group = money_raised_type)) +
  scale_color_manual(values = color_scale_0601) +
  geom_line() +
  geom_point(data = data_0601 %>% filter(money_raised_type == "progress_to_date" & days_since_launch == 10)) +
  geom_point(data = data_0601 %>% filter(money_raised_type == "last_year" & days_since_launch == 30)) +
  scale_y_continuous(limits = c(0, 60000), breaks = seq(0, 60000, 10000), labels = label_dollar()) +
  scale_x_continuous(breaks = seq(0, 30, 5), expand = expansion(add = c(.1, 3))) +
  # geom_hline(yintercept = 50000, color = GRAY) +
  geom_linerange(y = 50000, xmin = 0, xmax = 30, color = GRAY, size = .2) +
  annotate("text", x = 10.25, y = 39000, label = "Progress to date", color = BLUEGRAY, size = 4.5) +
  annotate("text", x = 31.7, y = 54800, label = "Last year", color = LIGHTSTEELBLUE, size = 3) +
  annotation_custom(progress_grob_0601, xmin = -10.1, ymin = 5000) +
  annotation_custom(last_year_grob_0601, xmin = 29, ymin = 40000) +
  geom_label(x = 2, y = 50000, label = "GOAL", fill = "white", label.size = NA, color = GRAY, size = 3) +
  labs(
    title = "Annual giving campaign progress",
    x = "Days since campaign launch",
    y = "Money raised"
  ) +
  my_swd_theme() +
  theme(
    axis.ticks.y = element_line(size = .2),
    axis.ticks.x = element_line(size = .2),
    plot.title = element_text(hjust = .08)
  ) +
  coord_capped_cart(bottom = "right", left = "top")
```

## Practice Workbook Figures

### 7.1

```{r}
data_pract_0701_OG <- read_csv("data/fig_practice_0701_data.csv")
```

```{r}
data_pract_0701 <- data_pract_0701_OG %>%
  pivot_longer(cols = `2015`:`2019`, names_to = "year", values_to = "percent") %>%
  mutate(
    percent = parse_number(percent),
    year = as_factor(year)
  )

blue_text_grob <- grobTree(richtext_grob(
  "<span style=\"fontsize: 1em\">The proportion of revenue from Feature <em>Z</em> is higher sooner:<br><span style=\"color:DARKBLUE\">28% of revenue was generated through Feature <em>Z</em> within<br> the first quarter advertising for new advertisers in 2019<br></span> compared to 15% for those creating<br> their first ad in 2015.</span>",
  gp = gpar(fontsize = 8), hjust = 0, align_widths = TRUE, padding = unit(c(6, 6, 4, 6), "pt")
))

green_text_grob <- grobTree(richtext_grob(
  "<span style=\"fontsize: 1em\"><br><br>The proportion of revenue from Feature <em>Z</em> generally<br>increases with advertiser tenure. <span style=\"color:FORESTGREEN\">By 16 quarters of<br>advertising, 50% of revenue came from Feature <em>Z</em><br>for those first creating an ad in 2015.</span> Given the<br>increase in sophistication over time, this may end<br>up being even higher for more recent cohorts.<br><br><br><br><br>We will continue to monitor and<br>report back if things change.</span>",
  gp = gpar(fontsize = 8), hjust = 1, align_widths = TRUE, box_gp = gpar(col = LIGHTSTEELBLUE, fill = NA, lwd = 2), padding = unit(c(4, 4, 2, 4), "pt")
))
```

Graph

```{r}
ggplot(data_pract_0701, aes(Age, percent, group = year)) +
  geom_line(color = GRAY, size = .8) +
  geom_point(color = GRAY, size = 3, data = data_pract_0701 %>% filter(Age == 0 & year != "2019")) +
  geom_point(color = DARKBLUE, size = 3, data = data_pract_0701 %>% filter(Age == 0 & year == "2019")) +
  geom_point(color = GREEN, size = 3, data = data_pract_0701 %>% filter(year == "2015" & Age == 15)) +
  scale_y_continuous(limits = c(0, 50), labels = label_percent(scale = 1), expand = expansion(add = c(1, 3))) +
  scale_x_continuous(breaks = seq(0, 16, 1), expand = expansion(add = c(.01, 2))) +
  geom_text(aes(label = year), color = GREEN, nudge_x = .8, data = data_pract_0701 %>% filter(year == "2015" & Age == 15)) +
  geom_text(aes(x = 11.75, y = 43.8, label = year), color = GRAY, data = data_pract_0701 %>% filter(year == "2016" & Age == 12)) +
  geom_text(aes(x = 8.55, y = 40, label = year), color = GRAY, data = data_pract_0701 %>% filter(year == "2017" & Age == 12)) +
  geom_text(aes(x = 4.6, y = 34.6, label = year), color = GRAY, data = data_pract_0701 %>% filter(year == "2018" & Age == 12)) +
  geom_text(aes(x = 1, y = 33, label = year), color = DARKBLUE, data = data_pract_0701 %>% filter(year == "2019" & Age == 12)) +
  annotation_custom(blue_text_grob, xmin = -20.2, ymin = 33) +
  annotate("text", label = "bold(Sophistication~increases~with~time)", x = 4.5, y = 50, parse = TRUE) +
  annotate("text", label = "bold(Sophistication~increases~with~age)", hjust = 0.45, x = 13.4, y = 29, parse = TRUE) +
  annotation_custom(green_text_grob, xmin = 16, ymin = -21) +
  annotate("text", label = "bold(No~action~is~needed~at~this~time)", x = 14.1, y = 7, parse = TRUE) +
  annotate("text", label = paste("FIRST AD", "CREATED", sep = "\n"), size = 2.5, hjust = 1, x = 18.2, y = 50, color = GRAY) +
  labs(
    title = "Feature Z % of revenue by advertiser age",
    x = "ADVERTISER AGE (QTRS)",
    y = "FEATURE Z % OF TOTAL REVENUE"
  ) +
  my_swd_theme() +
  theme(
    axis.title.y = element_text(size = 9),
    axis.title.x = element_text(size = 9)
  ) +
  coord_capped_cart(bottom = "right", left = "top")
```

### Case study #5

```{r}
pilot_data_OG <- read_csv("data/pilot_program_data.csv")
```

```{r}
pilot_data_2 <- pilot_data_OG %>%
  pivot_longer(-program_type, names_to = "item_type", values_to = "item_value") %>%
  mutate(
    program_type = str_to_upper(program_type)
  ) %>%
  mutate(
    program_type = as_factor(program_type),
    item_value = parse_number(item_value),
    item_type = fct_rev(as_factor(item_type))
  )

pilot2_color_scale <- c(
  "Bored" = GRAY,
  "Not Great" = GRAY,
  "OK" = ORANGE,
  "Kind of interested" = STEELBLUE,
  "Excited" = STEELBLUE
)

pilot_left_grob <- grobTree(richtext_grob(
  "<span style=\"fontsize:5rem; color:DIMGRAY\">BEFORE program, the<br><span style=\"color:ORANGE\">majority of children (40%)<br>felt just *OK*</span> about science.</span>",
  hjust = 0
))

pilot_right_grob <- grobTree(richtext_grob(
  "<span style=\"fontsize:5rem; color:DIMGRAY\">AFTER program, <span style=\"color:STEELBLUE\">more children<br>were *Kind of interested (30%)*<br>& *Excited (38%)*</span> about science.</span>",
  hjust = 1
))
```

Graph

```{r}
ggplot(pilot_data_2, aes(x = fct_rev(program_type), y = item_value, fill = item_type)) +
  geom_col(color = "white", width = .6) +
  scale_fill_manual(values = pilot2_color_scale) +
  scale_y_continuous(position = "right", labels = label_percent(scale = 1)) +
  scale_x_discrete(expand = expansion(add = c(1, .4))) +
  coord_flip() +
  my_swd_theme() +
  theme(
    axis.title.x = element_blank(),
    axis.title.y = element_blank(),
    axis.line.y = element_blank(),
    axis.ticks.y = element_blank(),
    plot.subtitle = element_markdown(hjust = .21, size = 12, padding = unit(c(.5, 0, .5, 0), "cm"))
  ) +
  labs(
    title = "How do you feel about science?",
    subtitle = paste0(
      "<span style='color:", pilot2_color_scale[1], "'>", "**Bored**", "</span>", " | ",
      "<span style='color:", pilot2_color_scale[2], "'>", "**Not Great**", "</span>", " | ",
      "<span style='color:", pilot2_color_scale[3], "'>", "**OK**", "</span>", " | ",
      "<span style='color:", pilot2_color_scale[4], "'>", "**Kind of interested**", "</span>", " | ",
      "<span style='color:", pilot2_color_scale[5], "'>", "**Excited**", "</span>"
    )
  ) +
  annotation_custom(pilot_left_grob, xmax = .6, ymax = 5) +
  annotation_custom(pilot_right_grob, xmax = .6, ymax = 205)
```

## Random
```{r}
div_OG <- read_csv("data/hr_diversity.csv")
```

```{r}
div_OG %>%
  select(Department, Salary, RaceDesc)  %>% 
  group_by(Department) %>% 
  count(RaceDesc)
```

```{r}
diversity_colors <- c(
  "White" = BLUEGRAY,
  "Black or African American" = LIGHTSTEELBLUE,
  "Two or more races" = STEELBLUE,
  "Asian" = GRAY,
  "American Indian or Alaska Native" = SILVER,
  "Hispanic" = NAVY
)

perc_group <- grobTree(richtext_grob(
  "<b>$51,400</b>",
  gp = gpar(col = LIGHTSTEELBLUE, fontsize = 8)
))
```
```{r}
div_OG %>%
  select(Department, Salary, RaceDesc)  %>% 
  mutate(across(where(is.character), as_factor)) %>% 
  group_by(Department, RaceDesc) %>% 
  summarise(n = n()) %>% 
  mutate(
    RaceDesc = fct_rev(RaceDesc),
    perc = round(n / sum(n), digits = 2)
  ) %>% 
  ggplot() +
    geom_col(mapping = aes(x = perc, y = Department, fill = RaceDesc), color = "white", width = .9) + 
    # scale_fill_brewer(palette = "Paired")
    scale_fill_manual(values = diversity_colors) +
    scale_x_continuous(position = "top", expand = expansion(add = c(0, .05)), labels = label_percent()) +
    my_swd_theme() + theme(
      legend.position = "top",
      legend.justification = c(0,1),
      legend.title = element_blank(),
      legend.text = element_text(size = 13),
      axis.title.x = element_blank(),
      axis.title.y = element_blank(),
      axis.line.y = element_blank(),
      axis.ticks.y = element_blank(),
      axis.text.y = element_text(size = 12),
      plot.title = element_text(margin = margin(1,1,15,1, unit = "pt"), size = 20),
      plot.background = element_rect(color = "white")
    ) +
    labs(
      title = "Distribution of employees by Department and Race",
    ) +
    coord_capped_cart(top = "both")

```