lubridate_livecode_solutions.Rmd

---
title: "Using lubridate to work with time intervals"
output: html_notebook
---

# This tutorial
When we analyze Electronic Health Records, we often want to know when a particular event took place within a patient's timeline to help us determine if it might have an impact on an health outcome of interest. 

The **lubridate** package can help us work with dates and times in our dataset!

# The data
For this tutorial, we'll use a dataset on Nobel Laureates originally prepared by [Kaggle](https://www.kaggle.com/nobelfoundation/nobel-laureates#archive.csv) and made available for [TidyTuesday on 2019-05-14](https://github.com/rfordatascience/tidytuesday/blob/3040e14f3b4934e6fc621f57d78324b53e549086/data/2019/2019-05-14/readme.md).

This dataset contains a few date variables for each laureate and will serve as a useful analogy for working with patient health records.

The dataset doesn't include the month and day that the awards were given but we know the awards are given on [December 10th every year, on the anniversary of Alfred Nobel's death](https://www.nobelprize.org/nobel-prize-award-ceremonies/).

We can use the `prize_year` and lubridate's `make_date()` function to create a date, and then use lubridate's `update()` function to update the date to reflect the month (12) and day (10) we need.

# Setup
```{r, message=FALSE}
# loading libraries ----
library(tidyverse)
library(lubridate)

# importing data ----
nobel_laureates <- readr::read_csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2019/2019-05-14/nobel_winners.csv")
nobel_laureates

# creating prize_date variable ----
nobel_laureates_prize_date <-
nobel_laureates %>%
  select(prize_year, full_name, gender, birth_date, death_date) %>%
  mutate(prize_year_new = make_date(prize_year)) %>%
  mutate(prize_date = update(prize_year_new, month = 12, mday = 10)) %>%
  select(-prize_year)
nobel_laureates_prize_date
```

# Example 1: Finding `age_at_award`

For this step we'll use lubridate's `interval()` and `years()` functions to find each laureate's age at the time they received the Nobel prize.

For more information: https://r4ds.had.co.nz/dates-and-times.html#intervals

```{r}
nobel_laureates_age <-
nobel_laureates_prize_date %>%
  mutate(age_at_award = interval(birth_date, prize_date)/years())
```

# Example 2: Creating `prize_in_lifetime`

For this step we'll use `%within%` from lubridate to check if the Nobel prize was awarded within each laureate's lifetime.

For more information: https://lubridate.tidyverse.org/reference/within-interval.html

```{r}
nobel_laureates_prize_in_lifetime <-
nobel_laureates_age %>%
  select(prize_date, everything()) %>%
  mutate(prize_in_lifetime = 
           ifelse(prize_date %within% interval(birth_date, death_date), "yes", "no")) 
```

----

# Plotting `age_at_award` categorized by `prize_in_lifetime` and by `gender`

For extra practice, we can plot the laureate's age at the time of the award using ggplot2's `geom_boxplot()` geometry, and look for differences according to:
- whether they received the Nobel prize in their lifetime
- gender (note: this variable is likely to be representative of sex assigned at birth rather than self-identified gender)

```{r}
nobel_laureates_prize_in_lifetime %>%
  filter(!is.na(prize_in_lifetime)) %>%
  ggplot(aes(x = prize_in_lifetime, y = age_at_award)) + 
  geom_boxplot(aes(fill = gender))
```