2021_02_02_hbcu.Rmd

---
title: "TidyTemplate"
date: 2021-02-02
output: html_output
---

# TidyTuesday

Join the R4DS Online Learning Community in the weekly #TidyTuesday event!
Every week we post a raw dataset, a chart or article related to that dataset, and ask you to explore the data.
While the dataset will be “tamed”, it will not always be tidy! As such you might need to apply various R for Data Science techniques to wrangle the data into a true tidy format.
The goal of TidyTuesday is to apply your R skills, get feedback, explore other’s work, and connect with the greater #RStats community!
As such we encourage everyone of all skills to participate!

```{r setup, include=FALSE}

knitr::opts_chunk$set(echo = TRUE)

library(tidyverse)
library(tidytuesdayR)
theme_set(theme_light())
library(scales)

```

# Load the weekly Data

Dowload the weekly data and make available in the `tt` object.

```{r Load}

tt <- tt_load("2021-02-02")

hbcu_all_long <- tt$hbcu_all %>%
  gather(metric, enrollment, -Year) %>%
  rename(year = Year)

hbcu_all_long %>%
  filter(str_detect(metric, " - ")) %>%
  separate(metric, c("degree_length", "type"), sep = " - ") %>%
  filter(degree_length != "Total") %>%
  ggplot(aes(year, enrollment, color = type)) +
  geom_line() +
  facet_wrap(~ degree_length) +
  labs(y = "# enrolled in HBCU",
       color = "")

hbcu_all_long %>%
  filter(metric %in% c("Males", "Females")) %>%
  ggplot(aes(year, enrollment, color = metric)) +
  geom_line() +
  expand_limits(y = 0) +
  labs(y = "# enrolled in HBCU",
       color = "")
```

* Most degrees from HBCU are 4-year, and 2-year degrees are almost entirely from public schools
* More women than men enroll in HBCU, and that has been increasing over time since 1980s

```{r}
hbcu_black_long <- tt$hbcu_black %>%
  gather(metric, black_enrollment, -Year) %>%
  rename(year = Year)

hbcu_compare_long <- hbcu_all_long %>%
  full_join(hbcu_black_long, by = c("year", "metric")) %>%
  mutate(pct_black = black_enrollment / enrollment)

hbcu_compare_long %>%
  filter(metric == "Total enrollment") %>%
  ggplot(aes(year, pct_black)) +
  geom_line() +
  scale_y_continuous(labels = percent) +
  expand_limits(y = 0) +
  labs(y = "% of HBCU enrollment that is Black")


hbcu_compare_long %>%
  filter(metric %in% c("Males", "Females")) %>%
  ggplot(aes(year, pct_black, color = metric)) +
  geom_line() +
  scale_y_continuous(labels = percent) +
  expand_limits(y = 0) +
  labs(y = "% of HBCU enrollment that is Black")

hbcu_compare_long %>%
  filter(str_detect(metric, "Total -")) %>%
  mutate(metric = str_remove(metric, "Total - ")) %>%
  ggplot(aes(year, pct_black, color = metric)) +
  geom_line() +
  scale_y_continuous(labels = percent) +
  expand_limits(y = 0) +
  labs(y = "% of HBCU enrollment that is Black",
       color = "")
```

```{r}
gather_race_ethnicity <- function(tbl) {
  tbl %>%
    mutate_if(is.character, parse_number) %>%
    rename(year = Total) %>%
    filter(!is.na(year)) %>%
    gather(race_ethnicity, value, -year) %>%
    mutate(column = ifelse(str_detect(race_ethnicity, "Standard Errors - "), "standard_error", "percent"),
           race_ethnicity = str_remove(race_ethnicity, "Standard Errors - ")) %>%
    spread(column, value) %>%
    mutate(standard_error = abs(standard_error)) %>%
    filter(!is.na(percent)) %>%
    mutate(race_ethnicity = str_remove(race_ethnicity, "1$"),
           percent = percent / 100,
           standard_error = standard_error / 100)
}

hs_over_time <- tt$hs_students %>%
  slice(-(1:3)) %>%
  gather_race_ethnicity()

bach_over_time <- tt$bach_students %>%
  gather_race_ethnicity()

education_over_time <- bind_rows(hs_over_time %>% mutate(degree = "High School"),
                                 bach_over_time %>% mutate(degree = "Bachelor's"))

hs_over_time %>%
  mutate(race_ethnicity = fct_reorder(race_ethnicity, -percent)) %>%
  ggplot(aes(year, percent, color = race_ethnicity)) +
  geom_line() +
  scale_y_continuous(labels = percent) +
  labs(color = "Race/ethnicity",
       y = "% of people aged >=25 who graduated HS") +
  expand_limits(y = 0)

bach_over_time %>%
  mutate(race_ethnicity = fct_reorder(race_ethnicity, -percent)) %>%
  ggplot(aes(year, percent, color = race_ethnicity)) +
  geom_line() +
  scale_y_continuous(labels = percent) +
  labs(color = "Race/ethnicity",
       y = "% of people aged >=25 who graduated a bachelor's program") +
  expand_limits(y = 0)

education_over_time %>%
  filter(year >= 1940,
         !str_detect(race_ethnicity, "Islander -")) %>%
  mutate(degree = fct_relevel(degree, "High School"),
         race_ethnicity = str_remove(race_ethnicity, "Total - ")) %>%
  mutate(race_ethnicity = fct_reorder(race_ethnicity, percent, last, .desc = TRUE)) %>%
  ggplot(aes(year, percent, color = race_ethnicity)) +
  geom_line() +
  facet_wrap(~ degree) +
  scale_y_continuous(labels = percent) +
  labs(x = "Year",
       color = "Race/ethnicity",
       y = "% of people aged >=25 who have this degree") +
  expand_limits(y = 0)
```

Bring in a new dataset on fields

```{r}
a25 <- readxl::read_excel("~/Downloads/A-25.xls")

a25_cleaned <- a25 %>%
  select(-starts_with("...")) %>%
  rename(field_gender = 1) %>%
  mutate(group = cumsum(is.na(field_gender))) %>%
  filter(!is.na(field_gender)) %>%
  select(group, everything()) %>%
  mutate(field_gender = str_remove(field_gender, " \\.\\.\\..*")) %>%
  group_by(group) %>%
  mutate(field = first(field_gender),
         gender = ifelse(field_gender %in% c("Men", "Women"), field_gender, "Total")) %>%
  ungroup() %>%
  select(field, gender, everything()) %>%
  select(-field_gender, -group)

a25_cleaned %>%
  select(field, gender, contains("HBCU")) %>%
  rename(pct_hbcu_total = 3,
         pct_hbcu_black = 4) %>%
  filter(gender != "Total") %>%
  mutate(field = fct_reorder(field, pct_hbcu_black, na.rm = TRUE),
         pct_hbcu_black = pct_hbcu_black / 100) %>%
  ggplot(aes(pct_hbcu_black, field, fill = gender)) +
  geom_col(position = "dodge") +
  scale_x_continuous(labels = percent) +
  labs(x = "% of first degrees from an HBCU, among Black students",
       y = "Field",
       fill = "")
```