# Exploratory RNAseq data analysis using RMarkdown - Day 10
Materials: [https://angus.readthedocs.io/en/2017/rmarkdown_rnaseq.html](https://angus.readthedocs.io/en/2017/rmarkdown_rnaseq.html)

## Background and Motivation
RMarkdown is a great tool for reproducible research. This lesson will use data from [this paper](http://rnajournal.cshlp.org/content/22/6/839.long). Today we will learn how to integrate our results into our writing, which is good for writing papers which links code results and text based documents. This is also helpful for working with collaborators.

Additional (and fantastic!) RMarkdown tutorial from Marian: [https://rpubs.com/marschmi/RMarkdown](https://rpubs.com/marschmi/RMarkdown)

### Literate Programming

Programming that contains tables and discriptions. There are three steps:

1. The source document and code is separated and **parsed**.
2. The source coded is **executed** to return results.
3. The source code and narratives are **mixed**.

### On similarity to Jupyter Notebooks

RMarkdown and Jupyter notebooks have a similar result, though Jupyter notebooks are more dynamic (easily change-able). Marian recommends learning whichever is easier for you (and I agree!).

It is possible to use other languages witin RMarkdown, similar to Jupyter notebooks.

### RMarkdown Features
* Use RStudio to view of the RMarkdown file in html.
* Bibliography references are coded into file (such as [1])
* The file contains code which generates plots. (Having this code readily available is super convenient for future modifications)

### RMarkdown vs HTML

HTML Example:

  <section>
    <h1>Fresh Berry Salad Recipe</h1>
    <ul>
      <li>Blueberries</li>
      <li>Strawberries</li>
      <li>Blackberries</li>
      <li>Raspberries</li>
    </ul>
  </section>
</body>

Markdown Example:

# Fresh Berry Salad Recipe

* Blueberries
* Strawberries
* Blackberries  
* Raspberries 

Markdown is certainly easier to read! Additionally, you can make the following types of documents from RMarkdown:
* HTML files (this is a _very_ powerful tool!)
* Word documents
* PDFs in LaTeX
* Slides

### Steps in RMarkdown Workflow
1. Open `.Rmd` file with a YAML header.
2. Write content using Markdown.
3. Embed code in chunks or inline.
4. Render the document into your beautiful final product!

### Code Steps to Create RMarkdown Rendered Document
1. Create `.Rmd` file from steps above
2. Use knitr R package to integrate R code into RMarkdown documents
3. Give created `.md` file to Pandoc, which converts the `.md` file to other formats (HTML in this case).

## Tutorial Steps

After downloading data and changing the the correct working directory by following steps in the materials, create a new RMarkdown file using the new file button.

### Anatomy of an RMarkdown File
#### 1. YAML headers
Nested list structure that exists at top of documents. RStudio automatically generated this upon creation of new RMarkdown file. Output file can be modified within this code chunk.

#### 2. Narrative or Description of analysis
This section will be descriptions of your science.

#### 3. Code
There are two methods to embed code.
1. **Inline Code**: This can be used to generate values, such as a p value, instead of just typing the value. Inline code should be surrounded by one backtick.
2. **Code Chunks**: This allows for multiple lines of code or scripts to be embedded within the document. Code chunks should be surrounded by 3 backticks.

### Knit it! (Execute Code and Create Document)
When you are ready to combine your code and writing, including running the code, click the Knit button to integrate. This button is near the top of the RStudio console and looks like a ball of yarn with knitting needles. A new HTML file will be created (if you opted to create an HTML file in the YAML header) and open up with executed code.

#### Jargon Explanation
Knitting is the verb of rendering code from the `.Rmd` file, which is executing code and creating and output format. The output is an rendered document, an HTML file in the case of this tutorial.

### Changing Pane Layout

In order to view the code from your RMarkdown file and tabs, click the Viewer tab, which is next to Files, Plots, etc. If the View tab doesn't work, click on the gear symbol and click `Preview on window pane`.

### Labeling Code Chunks

Giving labels to code chunks allows for easy navigation through long documents.

### Caching

If you have a big figure and don't want to regenerate it every time you knit, you can cache your figure. Information on caching is available on [section 8 of this tutorial](https://rpubs.com/marschmi/RMarkdown).