• Similar to this
• Probably include as a lab lesson or supplemental material

https://rstudio.github.io/learnr/

• Ugh, we need to do this. Cover in the API module. How to easily unnest complex nested JSON files?

Add at least one day of coverage for spatial data analysis using R. Cover basic data structures and visualization methods. No spatial regression or complex stuff, stick to more basics like mapmaking

Jeffrey Arnold exercise solutions

• Relational data
• Factors

See Stat 545 for inspiration. Don't need all of this though.

With the tidyverse one

Demo and practice flexdashboards

Write a brief guide for how to ask a question on Github. Include

• Reproducible example
• Formatting code snippets
• Session info

Analysis of age as a factor in NASA astronaut selection and career landmarks. Includes reproducible dataset to predict which astronaut applicants are selected for NASA training.

• Basics of animation, zoom and pan, brushing
• Using plot.ly to convert ggplot2 objects to ggplotly() graphs

Need a shell tutorial

render.sh correctly converts Jupyter Notebooks to markdown, but does not include a document title. Makes rendered HTML file look funny. Do notebooks come with titles that can be auto-rendered or forced via command line, or do I need to convert the first chunk title to the document title?

Take way too long to finish all 4 in class

Define a format for how to write R scripts (headers mainly). Add to hw00_homework_guidelines.Rmd

manifestoR

Page has now changed format. Probably try and find an easier example

Handwritten Digit Recognition (CLASSIFICATION PROBLEM)

Demonstrate the capabilities of statistical learning for a multi-class problem. Integrate into cm012

manifestoR

Something is hinky in stat005_resampling.Rmd.

These should be nearly identical, not such a large variation. Why?

Make clear that narratives matter in submission

Put under additional resources

Based on reprex. Add to lab01 or sometime in the first week of course

https://datalab.stanford.edu/

Store in a new repo under uc-cfss. Linked to via https://uc-cfss.github.io/fall2016

Use Jenny's data from the repurrrsive package to write a practice lesson on parsing recursive lists in R

Worth covering? If so, stash in week 9

With the Datasaurus Dozen

1. Software setup (special Wednesday edition)
2. Intro to R basics
3. Graphing tips - online only (MLK day)
4. The shell
5. Tools and approaches for debugging code (more specifically, functions)
6. R Markdown - if I don't introduce it in the first 5 weeks. Probably a good thing
7. Distributed computing - create account to access distributed computing server, whatever platform I decide upon
8. Setting up API keys

Revise for each course module

Add POS taggers and other demo packages.

• openNLP

• Appear to automatically include cross-validation
• Are they actually using distributed resources? Do they run faster than traditional algorithm?

Rather than just software install support, structure it a la Jenny Bryan's intro to R basics. Need to cover:

• Working directory
• Workspace (.Rdata)
• RStudio projects
• Difference between .R scripts and .Rmd R Markdown documents

Need to incorporate units on

• relational data
• strings
• factors
• maybe dates too, but probably leave that for students to figure out on their own

Cannot cram this into a single module. How can I adjust the rest of the schedule to make it work?

• Drop distributed computing
• Shift reproducibility to a lab session

Other ideas???

Replace 1:10 syntax with seq(from = 1, to = 10) syntax - much clearer to students

Rewrite the lesson to use a unique app. Find something using more social scientific data

Add to intro day

In-class exercise idea for sentiment analysis: https://paulvanderlaken.com/2017/08/03/harry-plotter-celebrating-the-20-year-anniversary-with-tidytext-the-tidyverse-and-r/

Analyzing Emotions using Facial Expressions in Video with Microsoft AI and R

Aka scoped verbs. See here for some documentation. Make sense to incorporate into conditional execution in cm008?

Add section on using random forests/SVMs to classify documents using text features

Needs more social scientific data examples to practice skills, not just datasets of convenience

• Revise to only use R scripts the first three weeks of the course
• Introduce R Markdown documents in cm008 or cm010

Any relevance to text mining lessons?

https://juliasilge.com/blog/song-lyrics-across/

Shifted to a paid API with private keys. Not a good, simple example anymore. Need to find a replacement that is free