Data Visualization: Theory and Practice (in R)
Athit Kao, PhD
UCI Bioinformatics Support Group - May 23, 2018
whoami
Prerequisites
Fundamental awareness of R and spreadsheet software
No programming experience necessary
R and RStudio installed (for 2nd half)
whoareyou
Who uses spreadsheet software to visualize data?
What types of figures (plots, charts, graphs, etc.) do you make?
Who has made a figure for a poster, presentation, or manuscript?
How many people have programmed before? What languages?
Comparison of Data Visualization Methods
Spreadsheet Software
- Many options (e.g. Sheets, Calc, Numbers, etc.)
- Quick editing and prototyping
- WYSIWYG plots
- Manual editing/manipulation error-prone
- File size limits, e.g. Excel has limit of 1M rows by 16K columns
Data Formats and Workflows
Click drag.. click.. click click...
What can we visualize here?
Histogram, Frequency Plot, Density Plot
- One-dimensional, not typically used to compare multiple groups
- Area is proportional to frequency of variable
- E.g. comparison of horsepower between all, manual, and automatic vehicles
Box Plot, Box-and-whisker Plot
- One-dimensional, typically used to compare multiple groups
- Minimally shows: First/third quartile, median, min/max
- Histogram + box plot = violin plot
- E.g. comparison of fuel efficiency across engine cylinder count
Scatter Plot, Scattergram
- Two-dimensional, can be used to compare multiple groups
- Each group has each datapoint's x and y values plotted
- E.g. Visualize the effect of engine displacement to horsepower across engine cylinder count
Writing Script vs. Using Console
Pre-loaded Data Sets in R
- E.g. iris, mtcars, airquality, faithful, etc.
- https://stat.ethz.ch/R-manual/R-devel/library/datasets/html/00Index.html
Why are we typing out everything?...
- So you can make mistakes:
- “That was case sensitive?”
- “I forgot a parenthesis/bracket?”
- “That wasn't a period/comma/semi-colon?”
- We learn better from mistakes
- We can help each other out here
Exercise #1a: Histogram
Exercise #1b: Density Plot
Exercise #2: Box Plot
Exercise #3: Scatter Plot
Basic Aesthetics
Saving Visualizations
Summarize for me...
- What are some advantages R has over Excel for data visualization?
- Why is it important to have code define a visualization?
- Anything weird or counterintuitive?
- T/F: Shared data and code will output the same exact visualization
- T/F: Anything spreadsheet software can visualize, R can visualize better
- T/F: You need to pay for “Premium Enterprise R” to make violin plots
- #4, TRUE; #5, TRUE; #6, FALSE!!!
Questions?
- Just try new code and see what happens; this isn't wet lab
- Google: Don't just search with “R”, use “R language”
- Stack Overflow: Use tag “[r]”
Keep going, you got this!
App: The Assignment Operator
- Use “<-” or “=”?
- I use “=” for consistency (with other programming languages) with no issues
- Consistency even within R
App: Tilde Operator ("~")
- Operator for use in “formulas”, used in many different functions and packages
- E.g. “y ~ x” means y (or whatever is on the left) is the response to x (or whatever is on the right)
- E.g. “mpg ~ cyl + am” means mpg depends on cylinder count and whether it's automatic transmission
App: NIH Proficiency Scale
- 1. Fundamental Awareness (basic knowledge): Common knowledge/understanding of basic techniques/concepts
- 2. Novice (limited experience): Expected to need help when performing this skill
- 3. Intermediate (practical application): Able to successfully complete tasks; expert required occasionally
- 4. Advanced (applied theory): Able to successfully complete tasks without assistance
- 5. Expert (recognized authority): Can provide guidance, troubleshooting, and answers related to this skill
- Source: https://hr.nih.gov/working-nih/competencies/competencies-proficiency-scale