2_qplot.rmd

---
title: |-
  R Crash Course \
   Part 2 -- *qplot()*
author: "Rob Colautti"
---

# 1. Getting Started
Install the `ggplot2` package

```{r}
library(ggplot2)
```

The *ggplot2* package includes two main functions, quickplot `qplot()` for fast graphs and the `ggplot()` function for more detailed, customizable graphs.

The package was created by the same group that also made:

*RStudio*, *RMarkdown*, *Shiny*, *ggvis*

and many other cool packages. These are available on the RStudio website:

https://www.rstudio.com/products/rpackages/

### Full ggplot2 Documentation

http://docs.ggplot2.org/current/

This tutorial focuses on ***qplot()***, a quick and simple yet versatile plotting function. It is much more intuitive than R's default graphics commands.

For more advanced graphics, see the [ggplot tutorial](./3_ggplot.html)

### General format:
`qplot(x=my.xVariable, y=my.yVariable, data=my.data.frame)`

## ppt slides
  1. Graphical concepts

  2. ggplot grammar
  
  3. Explanation of selection meta-analysis data

  4. anatomy of a graph
  
  [Slides](Graphics_small.pdf)
  
## Data setup

We will again be working with the FallopiaData.csv dataset, which can be [downloaded here](FallopiaData.csv), and saved to your project folder to follow along.

```{r}
MyData<-read.csv("FallopiaData.csv",header=T)
```

<br>

***

<br>

# 3. Basic Graphs
## One continuous
#### Produces a histogram by default
```{r}
qplot(x=Total,data=MyData)
```

## One categorical
#### Bar graph of counts for each category
```{r}
qplot(x=Scenario,data=MyData)
```

## Two continuous
#### Bivariate plot
```{r}
qplot(x=Silene,y=Total,data=MyData)
```

## Categorical by continuous
#### Categorical scatter plot
```{r}
qplot(x=Nutrients,y=Total,data=MyData)
```

<br>

***

<br>

# 4. Basic customization

### *binwidth =*
#### Change bin width of a histogram
```{r}
qplot(x=Total,data=MyData,binwidth=9)
qplot(x=Total,data=MyData,binwidth=0.5)
```

### *size =*
#### Controls size of points, lines, etc.
Google "pch shapes in R"
http://vis.supstat.com/2013/04/plotting-symbols-and-color-palettes/
```{r}
qplot(x=Silene,y=Fallopia,data=MyData,size=Total) # Scale by a variable
qplot(x=Silene,y=Total,data=MyData,size=I(5)) # Scale by a constant
```
**NOTE:** use of identity function: ***I()*** for constant

Compare to:
```{r}
qplot(x=Silene,y=Total,data=MyData,size=5)
```
Without I(), 5 is interpreted as a factor

### *colour =* or *color =*
#### Colours points based on a factor or 
Again, note use of ***I()*** for constants vs variables
```{r}
qplot(x=Silene,y=Fallopia,data=MyData,colour=Nutrients) # Categorical colour
qplot(x=Silene,y=Fallopia,data=MyData,colour=Total) # Continuous colour
qplot(x=Silene,y=Total,data=MyData,colour=I("purple")) # basic colour
qplot(x=Silene,y=Total,data=MyData,colour=I(rgb(1,0,0))) # rgb = red/green/blue
```

#### Note for histograms:
Colour applies to outlines, not so helpful
```{r}
qplot(x=Total,data=MyData,group=Nutrients,colour=Nutrients)
```

### *fill =* 
#### More useful for histograms
```{r}
qplot(x=Total,data=MyData,group=Nutrients,fill=Nutrients)
```

### *posit =* 
#### Avoid stacking histogram bars
```{r}
qplot(x=Total,data=MyData,group=Nutrients,fill=Nutrients)
qplot(x=Total,data=MyData,group=Nutrients,fill=Nutrients,posit="dodge")
```

### *alpha =*
#### Adjust transparency for overlapping points
```{r}
qplot(x=Silene,y=Total,data=MyData,size=I(5),alpha=I(0.3))
```

### *shape =*
#### Use different shapes 
Most common shape codes: 0-25
```{r}
qplot(x=Silene,y=Total,data=MyData,size=I(5),shape=Nutrients)
qplot(x=Silene,y=Total,data=MyData,size=I(5),shape=I(17))
```

#### Some shape types have outlines
You can use ***fill =*** and ***colour =*** to customize
```{r}
qplot(x=Silene,y=Total,data=MyData,size=I(5),shape=I(21),fill=I("yellow"),colour=I("red"))
```

### *xlab =*  and   *ylab =*
#### Custom axis labels
```{r}
qplot(x=Silene,y=Total,data=MyData,xlab="Silene Biomass",ylab="Total Biomass")
```

### *main =*
#### Add a title
```{r}
qplot(x=Silene,y=Total,data=MyData,main="Biomass")
```

<br>

***

<br>

# 5. Changing themes & Geoms
### ***+ theme_NAME()***
#### Modify basic format and appearance
These can be completely customized, but there are several pre-cut options:
```{r}
qplot(x=Silene,data=MyData) + theme_grey() # default
qplot(x=Silene,data=MyData) + theme_bw() # cleaner, better contrast
qplot(x=Silene,data=MyData) + theme_linedraw() # thicker grid lines
qplot(x=Silene,data=MyData) + theme_light() # fainter border and axis values
qplot(x=Silene,data=MyData) + theme_minimal() # no borders
qplot(x=Silene,data=MyData) + theme_classic() # x and y lines only, no tick marks
```

### Or create your own custom theme:
#### Simplified/clean version of theme_classic with bigger axis labels more suitable for presentation or publication:
```{r}
# Clean theme for presentations & publications used in the Colautti Lab
theme_pubworthy <- function (base_size = 12, base_family = "") {
  theme_classic(base_size = base_size, base_family = base_family) %+replace% 
    theme(
      axis.text = element_text(colour = "black"),
      axis.title.x = element_text(size=18),
      axis.text.x = element_text(size=12),
      axis.title.y = element_text(size=18,angle=90),
      axis.text.y = element_text(size=12),
      axis.ticks = element_blank(), 
      panel.background = element_rect(fill="white"),
      panel.border = element_blank(),
      plot.title=element_text(face="bold", size=24),
      legend.position="none"
    ) 
}
```

#### What it looks like:
```{r}
qplot(x=Silene,data=MyData) + theme_pubworthy() # A clean format with bigger axis labels for publication
qplot(x=Pct_Fallopia,y=Silene,data=MyData) + theme_pubworthy() # Bivariate plot
```

### ***geom =***
See website for list of geoms

http://docs.ggplot2.org/

A couple of examples:
```{r}
qplot(x=Nutrients,y=Total,data=MyData) # Basic
qplot(x=Nutrients,y=Total,data=MyData,geom="boxplot") # Geom
qplot(Total,data=MyData) # Basic
qplot(Total,data=MyData,geom="density") # Geom
```

<br>

***

<br>

# 6. Multiple graphs

### ***facets = ***
#### General form: facets=Vertical~Horizontal
```{r}
qplot(x=Silene,data=MyData,facets=Nutrients~.) # Vertical stacking
qplot(x=Silene,data=MyData,facets=.~Nutrients) # Horizontal stacking
qplot(x=Silene,data=MyData,facets=Taxon~Nutrients) # Both (2 column stacking)
```

<br>

***

<br>

# 7. Save output
#### Saving graphs/images requires three steps

  1. Open a file (e.g. pdf, svg, png)

  2. Create the graph inside the file

  3. Close the file 
      + Important: If you don't close the file, it is unusable.

```{r,eval=F}
pdf("SileneHist.pdf") # 1. Create and open a file
  qplot(x=Silene,data=MyData,facets=Taxon~Nutrients) # 2. Write the graph info
dev.off() # 3. Close the file
```
Note how the qplot command does not open in the plots window

This is because the info is sent to "SileneHist.pdf" instead