Reproducible Research in R

class: center, middle, inverse, title-slide

# Reproducible Research in <code>R</code>
## Lesson 2: Reading and Writing Data
### Christy Garcia and Chris Prener
### September 19, 2018

---

# Getting Started

1. Go to https://github.com/slu-dss/research-02/

2. Scroll down to "Lesson Quick Start"

3. Copy and paste the `install.packages()` function into your console and run

4. Copy and paste the `usethis::use_course()` function into your console and run

---

# Upcoming Events

### DSS 06: Reproducible Research in `R`

* Session 03 - October 3rd
* Session 04 - October 17th

### Other DSS Events

* Open Science Meetup - October 5th
* Brownbag, Power Analyses in `R` (South Campus) - October 31st
* Brownbag, Using Python for Automating Routine Tasks - November 28th

### RLadies

* Reproducible Research in `R` - October 3rd
* Plotting something great: Making graphs in `R` - November 14th

---

# Quick Review

1. `R` stores information in *objects* that are saved in your *enviornment*

2. All objects have *names*:

* names should not have spaces or wonky special characters
    
    * names should be short and descriptive
    
    * names should use either `dot.case`, `snake_case`, or `camelCase`
    
3. We assign data to objects with the assignment operator: `df <- data`

4. We interact with `R` using *functions*

5. Functions typically take one or more *arguments* as input

6. Functions can be *nested* within other functions

---

# File Paths

Every file on your computer has an "address" that locates it:

`/Users/chris/Documents/DSS/research-02/README.md`

`C:\Users\chris\Documents\DSS\research-02\README.md`

### Three problems

1. We are so used to using Windows File Explorer and macOS Finder to find files, or searching, that constructing valid file paths is no longer a typical skill for the median user. *File paths are a lost art.*

2. Depending on how elaborate your organizational structure is, this can consist of many different "layers" of folders. *File paths are a pain to type.*

3. File paths are typically specific to a user, so if our notebooks used hard-coded file paths and we gave the files to a colleague to execute, they would have to edit each file path call to match the specifics of their machines. *Fil paths limit reproducibility.*

---

# The Working Directory

To solve the problems of file paths, we use "working directories" - a starting place that we can associate with relative paths.

```r
> getwd()
[1] "/Users/chris/Documents/DSS/research-02"
```

Once we have a working directory set, we can instruct `R` using relative file paths:

```r
df <- readr::read_csv(file = "data.csv")
```

`R` will assume that the file `data.csv` sits within

`/Users/chris/Documents/DSS/research-02`.

---

# The Working Directory

To solve the problems of file paths, we use "working directories" - a starting place that we can associate with relative paths.

```r
> getwd()
[1] "/Users/chris/Documents/DSS/research-02"
```

Once we have a working directory set, we can instruct `R` using relative file paths:

```r
df <- readr::read_csv(file = "data/data.csv")
```

`R` will assume that the file `data.csv` sits within

`/Users/chris/Documents/DSS/research-02/data`.

---

# Setting the Working Directory Dynamically

To ensure that `R` code executes in *any* given environment without having to manually enter the working directory, we will use two tools:

1. `R` projects (e.g. `research-02.Rproj`)
2. The `here` package and its `here()` function

### R projects
`R` projects, when utilized with RStudio, automatically set the working directory to the directory where the `.Rproj` file is located.

If the following exists:

`/Users/chris/Documents/DSS/research-02/research-02.Rproj`

then

`/Users/chris/Documents/DSS/research-02`

is the working directory.

---

# Setting the Working Directory Dynamically

To ensure that `R` code executes in *any* given environment without having to manually enter the working directory, we will use two tools:

1. `R` projects (e.g. `research-02.Rproj`)
2. The `here` package and its `here()` function

### here
The `here` package allows us to both read and write to *subdirectories* of our working directory cleanly:

If we are using a `.Rproj` *and* the following exists:

`/Users/chris/Documents/DSS/research-02/data/data.csv`

then we will write:

```r
df <- readr::read_csv(file = here("data", "data.csv"))
```