STA/OPR 9750 - Week 8

Michael Weylandt

STA/OPR 9750 Mini-Project #02

Submission due yesterday at 11:45pm

Y’all have it out for The Shawshank Redemption…

  • Very entertained by fake posters
  • Impressive graphics and analysis - very nice.

The Shark!

STA/OPR 9750 Mini-Project #02

Add the following to your .gitignore file:

/.quarto/
.Rproj.user
*/*_cache/*
*tsv*
*csv*
*xlsx*
*zip
*gz
.DS_Store
.Rhistory

This instructs git to ignore matching files - less chance of accidental “super commits.”

(Possible to override if you necessary)

STA/OPR 9750 Mini-Project #02 - Peer Feedback

Peer feedback assigned on GitHub this morning

  • \(\approx 4\) feedbacks each
  • Take this seriously: around 20% of this assignment is “meta-review”
  • Goal: rigorous constructive critique

Submissions may not map perfectly to rubric - use your best judgement

Learn from this! What can you adapt for MP#03?

STA/OPR 9750 Mini-Project #03

Now online

Due November 13th

  • GitHub post (used for peer feedback) AND Brightspace

  • Three Weeks: don’t wait until the very end

  • Should be less demanding than MP #01 and MP#02

    • Lots of little files. No big files!
    • Maps and election retrodiction

Pay attention to the rubric

Thank You!

Thank you!

A personal note, if you allow me:

I’m really enjoying this class - thank you all!

Your effort is not unnoticed - I know this class starts “pedal-to-the-metal” but hopefully you’ve seen just how powerful these tools R.

More than that - I appreciate your good attitude and willingness to share your frustrations and triumphs. Reading comments on PA quiz this week was uplifting.

Continual Improvement

I’ve set up a TODO file with everything I want to improve for next cohort.

Suggestions welcome.

I will also ask Charles to distribute an opportunity for anonymous feedback.

Going Forward

Upcoming Mini-Projects

Tentative Topics

  • MP#04: Something financial

Any requests?

Course Project

Feedback from Charles and from me - I’m behind on returning.

Project should be your main focus for rest of course

  • Still need to do mini-projects and pre-assignments

Pre-Assignments

Brightspace - Wednesdays at 11:45

  • Reading, typically on course website
  • Brightspace auto-grades
    • I have to manually change to completion grading

Next pre-assignment is October 30th

Thank you for FAQs and (honest) team feedback. Keep it coming!

Course Support

  • Synchronous
    • Office Hours 4x / week
      • MW Office Hours on Monday + Thursday for rest of semester
      • CR Tuesday + Friday
      • No OH during Thanksgiving break
  • Asynchronous
    • Piazza (\(<50\) minute average response time)

Upcoming Week

Due Wednesday at 11:45pm:

  • Pre-Assignment #09 (Brightspace)
    • Data Import
  • MP #02 Peer Feedback on GitHub AND Brightspace

Pre-Assignment #08 FAQs

FAQ: ggplot2 - aes()

What is the aes function - stands between data and geom_

  • Each geom_ takes a fixed set of “coordinates”
  • Each data set has its own column names
  • aes ties these together
library(ggplot2); library(patchwork)
p1 <- ggplot(iris, aes(x=Sepal.Length, y=Sepal.Width)) + geom_point()
p2 <- ggplot(iris, aes(y=Sepal.Length, x=Sepal.Width)) + geom_point()
p1 + p2

FAQ: ggplot2 - Why do Pie Charts have a bad reputation?

  • Use of area and angle over length: less accurate perception
  • Depends on fill to convey category - limited categories

But honestly - “insider smugness” and hate of Excel

FAQ: ggplot2 - Plot Type Choice

For me:

  • Exploratory mode:
    • Simple: line, scatter, bar, frequency
  • Publication mode:
    • Very context specific

FAQ: ggplot2 - Font Sizing

Theme machinery!

ggplot(iris, aes(x=Sepal.Length, y=Sepal.Width)) + 
    geom_point() + theme(axis.text = element_text(size=24))

FAQ: ggplot2 - Overplotting / ScatterBlobs

Student asked about “scatterblobs” - typo(?) but I love it!

  • Density based plotting: hexbins, histograms, rugplots
  • Data reduction: summarization or sub-sampling

FAQ: ggplot2 - Optimizing Performance

Active project of ggplot2 team - not much you can do

Practical advice: plot less (see previous slide)

FAQ: ggplot2 - Beyond Scatter and Line

Some favorite semi-advanced plot types:

  • Violin plots: combination of boxplot and histogram
  • Ridgelines
  • Beeswarms

Deep rabbit hole

FAQ: ggplot2 - Geospatial Visualizations

That’s our goal for today!

FAQ: ggplot2 - High-Dimensional Data

High-dimensional data: measure many variables per observation (“wide”)

High-dimensional data is hard to visualize

Approaches:

  • Pair plots for “moderate” HDD
  • PCA (or similar dimension reduction. Take 9890!)

FAQ: ggplot2 - Creating a Custom Theme

my_theme <- theme_bw() + theme(panel.background = element_rect(fill = 'lightblue'))
ggplot(iris, aes(x=Sepal.Length, y = Sepal.Width)) + geom_point() + my_theme

Advanced:

  • theme_set() - change ggplot2 defaults
  • .Rprofile - set code to run every time you start R

FAQ: ggplot2 - When Not to Use

ggplot2 is designed to make good statistical graphics. Sub-par for:

  • Advanced interactivity
  • Really big data
  • Hardcore customization / “infographics”

FAQ: git WTF

Reference: Happy Git with R

Today

Warm-Up

“Datasaurus Dozen”:

  • install.packages("datasauRus") (Note capital R)
  • library(datasauRus); data(datasaurus_dozen)

Create an animated (gganimate) plot:

  • \(x, y\) scatterplot
  • Animate different values of dataset

If you are having trouble with gganimate, facet instead.

Warm-Up

Diving Deeper with ggplot2

Today: maps!

Install the sf package: Simple Features for Spatial Data

Exercise: Lab #08

Breakout Rooms

Room Team Room Team
1 Rat Pack 6 Ca$h VZ
2 Subway Surfers 7 Listing Legends
3 Chart Toppers 8 TDSSG
4 Metro Mindset 9 Broker T’s
5 Apple Watch 10 EVengers