data_cleaning.Rmd

---
title: "Data cleaning"
output: 
  learnr::tutorial:
    progressive: true
    allow_skip: true
runtime: shiny_prerendered
description: >
  Learn how to get started with R and RStudio, and how to import a dataset
---

<!-- Add JavaScript code for making the exercise code larger -->
<script language="JavaScript" src="js/exercise-font-size.js"></script>

```{r setup, include=FALSE}
# load packages ----------------------------------------------------------------
library(learnr)
library(gradethis)
library(tidyverse)
library(here)
library(rio)
library(basket) # not sure if we need this
library(etude) # helper functions for gradethis

# set options for exercises and checking ---------------------------------------
gradethis_setup()

learnr::tutorial_options(exercise.timelimit = 60) 
  #exercise.checker = gradethis::grade_learnr) 
    # alternatively, submitr::null_code_checker


# event recorder ---------------------------------------------------------------
# see https://github.com/dtkaplan/submitr/blob/master/R/make_a_recorder.R
tutorial_options(exercise.eval = FALSE)  # pre-evaluate exercises

vfun <- submitr::make_basic_validator(NULL, "hello") #basket::check_valid

new_recorder <- function(tutorial_id, tutorial_version, user_id, event, data) {
    cat(
      tutorial_id, 
      " (v", tutorial_version, "); ",
      format(Sys.time(), "%Y-%M%-%D %H:%M:%S %Z"), "; ",
      user_id, "; ",
      event, "; ",
      data$label, "; ",
      data$answers, "; ",
      data$code, "; ",
      data$correct, "\n", sep = "",
      
      file = here::here("event_records", "learnr_basics.txt"),
      append = TRUE)
}

options(tutorial.event_recorder = new_recorder)


# hide non-exercise code chunks ------------------------------------------------
knitr::opts_chunk$set(echo = FALSE)


# data prep --------------------------------------------------------------------
linelist_raw <- rio::import(here::here("data", "linelist_raw.xlsx"))
linelist <- rio::import(here::here("data", "linelist_cleaned.rds"))
malaria_counts <- rio::import(here::here("data", "malaria_facility_count_data.rds"))
```


```{r}
submitr::login_controls() # show login and password with "Submit" button.
```


```{r context = "server", echo = FALSE}
# see https://rdrr.io/github/dtkaplan/submitr/f/vignettes/using.Rmd
options(tutorial.storage = "none")
vfun <- submitr::make_basic_validator(NULL, "hello")       #basket::check_valid
storage_actions <- submitr::record_local("./minimal_submissions.csv")
submitr::shiny_logic(input, output, session, vfun,
                     storage_actions)
```




## Introduction to R for Applied Epidemiology and Public Health
### Data cleaning and core functions

```{r appliedepi-banner, fig.margin = TRUE, echo = FALSE, fig.width = 3, out.width = "100%", fig.cap = ""}
knitr::include_graphics("images/moz-banner.png")
```


### Welcome

Welcome to the course "Introduction to R for applied epidemiologists", offered for free by [Applied Epi](www.appliedepi.org) - a non-profit organisation that offers open-source tools, training, and support to frontline public health practitioners.

This interactive tutorial focuses on **cleaning of datasets often encountered by applied epidemiologists and public health practitioners**, such as outbreak linelists, surveillance, and laboratory data.  


#### Target Audience  

This course is designed with the following objectives: 

* To be friendly to people who have never used a programming language before
* To teach R emphasizing examples, datasets, and challenges commonly faced by applied epidemiologists
* To be modular - so that you can skip to section most relevant to you

If this is your first introduction to R programming, please consider first completing our [R Setup and Data Import tutorial], which introduces R, RStudio, R projects, R code syntax, and explains how to import a dataset into R.  



#### Other languages
This course is available...


#### Offline / Online

You can access this tutorial offline by downloading our R package ... 
If viewing offline, you can view the videos by doing ...



#### Learning goals

In this tutorial you will learn and practice:  

* Using the pipe operator (`%>%`) to pass the dataset from one cleaning function to another
* The core **tidyverse** R functions used to reduce, clean, and modify data frames  
* ...  
* ...  



This tutorial adapts the [Data cleaning and core functions](https://epirhandbook.com/en/cleaning-data-and-core-functions.html) section of our free [ Epidemiologist R Handbook](https://epirhandbook.com/en/), which is available for use offline as well. 


#### Data consent

We continually improve these tutorials by collecting your entries and submitted answers to the quiz questions. By continuing, you consent to this collection and use.

To continue anonymously... do XYZ.



#### Who made this course  

This course is designed by epidemiologists with decades of ground-level experience in outbreak response and local public health work. 


```{r appliedepi-hexes, fig.margin = TRUE, echo = FALSE, fig.width = 3, out.width = "50%", fig.cap = ""}
knitr::include_graphics("images/hex-sidebyside.png")
```


## Data used and directory structure

In this tutorial we will use the following datasets. Please take a few minutes to review the structure and content of each dataset before continuing.

Use the arrows on the right to scroll through hidden columns. Note that these are "raw" (messy) datasets that mimic problems commonly found in real-life epidemiological datasets.  


### **A "linelist" of cases in a fictional (not real) Ebola outbreak***

A "linelist" is a term used in applied epidemiology to refer to a table that contains key information about each case or suspect case in an outbreak. Each row represents one case, and the columns contain variables such as age, sex, date of symptom onset, outcomes, etc.

This dataset contains `r nrow(linelist_raw)` rows and `r ncol(linelist_raw)` columns. Below are the first 5 rows:  

```{r}
head(linelist_raw)
```

Click to [download the **raw** dataset](https://github.com/appliedepi/epirhandbook_eng/raw/master/data/case_linelists/linelist_raw.xlsx) for your own practice.

Click to [download the **clean** dataset as an **.rds file**](https://github.com/appliedepi/epirhandbook_eng/raw/master/data/case_linelists/linelist_cleaned.rds) for your own practice. A *.rds file* is an R-specific file type that preserves column classes. This ensures you will have only minimal cleaning to do after importing the data into R.

### **Aggregated data from malaria surveillance in a fictional country**  

Aggregated data in epidemiology usually means a table of counts for each facility, or district, etc. Sometimes, the counts can also be per day, week, or month.  

In this fictional dataset, each facility reported *daily* case counts of rapid-test (RDT)-confirmed malaria. Thus, each row represents the number of cases for a specific facility on a specific day.  

This dataset contains `r nrow(malaria_counts)` rows and `r ncol(malaria_counts)` columns. Below are the first 5 rows:  

```{r}
head(malaria_counts)
```


Click to [download the **clean** malaria counts dataset as an **.rds file**](https://github.com/appliedepi/epirhandbook_eng/raw/master/data/malaria_facility_count_data.rds) for your own practice. A *.rds file* is an R-specific file type that preserves column classes. This ensures you will have only minimal cleaning to do after importing the data into R.


### Directory structure

Photo or GIF of directory structure

### Accessing example data
Here is how to access the example data

### Tidy Data

We highly recommend doing our tutorial on Tidy Data in Applied Epidemiology. Collecting, formatting, and preparing your dataset *before* importing it into R is a critical step!
LINK
VIDEO TEASER




## Install and Load R packages {#packages}

To use basic functions with public health data, the tidyverse metapackage is very useful. Tidyverse loads the dplyr, ggplot2, and other packages that are useful in epi data analysis. 

We've preloaded the below packages for now. Installation and loading of these packages is described on the EpiRHandbook [_Suggested Packages_](https://epirhandbook.com/en/suggested-packages-1.html) page. 


In this assignment we'll work with X R packages, let's load them!

```{r load-package, exercise = TRUE}
pacman::p_load(___)
```

```{r load-package-hint}
pacman::p_load(rio, here, janitor, tidyverse)

```

```{r load-package-check}
grade_this_code("You are correct, the packages you need for this lesson are now loaded!")
```

```{r eval=FALSE}
pacman::p_load(learnr, here, rio, janitor, tidyverse)
```

### Recommended R packages for public health

See this Epi R Handbook LINK for our recommended packages.









## Import data {#import} 
https://www.epirhandbook.com/en/import-and-export.html
Import that data and save it as "raw" file

To import data from a sub-folder, the `import()` command should be modified so that it correctly tells R where to search for this file. This is done using the here() function. 

```{r import-demo-subfolder, echo=T, eval=F}
linelist_raw <- import(here("data", "linelist_raw.xlsx"))   # import data and save as named object
```


## Begin pipe chain

What is a pipe chain?
Ways to see how the pipe chain works

### Quiz


## Clean columns

What is a column? Why not call it a variable?

### Clean names
janitor::clean_names()

### Select columns
You can also re-name them with select()

### Quiz


## Clean rows

### Drop missing values
### Deduplication
Touch upon some of these: https://www.epirhandbook.com/en/de-duplication.html
### Filter by logic
### Quiz




## New columns
### mutate()
### Re-coding
### Quiz




## Dates
See Epi R Handbook chapter on working with dates
https://www.epirhandbook.com/en/working-with-dates.html


### Convert to date
#### Character to date
#### Excel to date

### epiweeks
use floor_date() from lubridate
#### ISO weeks

### Quiz


## Pivots
See Epi R Handbook page on Pivoting
https://www.epirhandbook.com/en/pivoting-data.html
### Quiz

## Joins
https://www.epirhandbook.com/en/joining-data.html
### Quiz

## Next steps


##  TEMPLATE Exercises

### Exercise with Code

*Here's an exercise with some prepopulated code as well as `exercise.lines = 5` to provide a bit more initial room to work.*

Now write a function that adds any two numbers and then call it:

```{r add-function, exercise=TRUE, exercise.lines = 5}
add <- function() {
  
}
```

### Exercise with Hint

*Here's an exercise where the chunk is pre-evaulated via the `exercise.eval` option (so the user can see the default output we'd like them to customize). We also add a "hint" to the correct solution via the chunk immediate below labeled `print-limit-hint`.*

Modify the following code to limit the number of rows printed to 5:

```{r print-limit, exercise=TRUE, exercise.eval=TRUE}
mtcars
```

```{r print-limit-hint}
head(mtcars)
```

### Quiz

*You can include any number of single or multiple choice questions as a quiz. Use the `question` function to define a question and the `quiz` function for grouping multiple questions together.*

Some questions to verify that you understand the purposes of various base and recommended R packages:

```{r quiz}
quiz(
  question("Which package contains functions for installing other R packages?",
    answer("base"),
    answer("tools"),
    answer("utils", correct = TRUE),
    answer("codetools")
  ),
  question("Which of the R packages listed below are used to create plots?",
    answer("lattice", correct = TRUE),
    answer("tools"),
    answer("stats"),
    answer("grid", correct = TRUE)
  )
)
```