Learning outcomes

By the end of this course, students will be able to...

Setup

Isolate
- data using filter (e-data01-isolate)
- variables using select (e-data01-isolate)
Recognize
- basic data objects: vectors, types (e-setup05-vectors, e-setup06-types)
Derive
- new variables with mutate (e-data02-derive)
- summaries with group_by and summarize (e-data02-derive)
- string matches with regular expressions (e-data06-strings)
- positional and ranking relations with window functions (e-data08-window)
Tidy
- by reshaping with pivot_longer and pivot_wider (e-data03-pivot)
- by separate and unite of columns (e-data04-separate-unite)
- by join of multiple tables (e-data05-join)
Load
- data from a flat file, e.g. a comma-separated value (csv) file (e-data00-basics)
- data from an Excel sheet (e-data09-readxl)
- data from googlesheets (e-data12-simple-pipeline)
Wrangle
- data from a messy source (e-data09-readxl, e-data13-cleaning)
- strings with stringr (e-data06-strings)
- complex code into readable, functional pipelines (e-data07-placeholders)
- loops into functional calls using purrr (e-data10-map)
- factors with forcats (e-data11-factors)
Liberate
- data from images with WebPlotDigitizer (e-data14-webplotdigitizer)
- data from printed tables with Tabula (e-data15-tabula)

State and use the basic components of the grammar of graphics:
- geometry (many!)
- aesthetics (e-vis00-basics)
- theme (e-vis06-themes)
- layers (e-vis04-scatterplot)
Create
- bar charts (e-vis01-bar-charts)
- histograms (e-vis02-histograms)
- boxplots (e-vis03-boxplots)
- scatterplots (e-vis04-scatterplot)
- line plots (e-vis05-lines)
- small multiple plots (e-vis06-multiples)
- tweaks to visual aesthetics (e-vis06-themes)
Design
- EDA-quality graphics (many!)
- presentation-ready graphics (e-vis06-themes, e-comm01-story-basics)
- comparisons using small multiples (e-vis08-multiples)
- improved visuals through an iterative approach (e-vis09-improve)
Criticize
- data visualizations based on the visual hierarchy (e-vis07-perception)

Distinguish
- between error and uncertainty (c02-michelson, e-stat07-error-bias)
- between populations and samples (e-stat04-population)
Describe
- a random quantity with a distribution (e-stat01-distributions)
- a dataset with descriptive statistics (e-stat03-descriptive, e-stat05-moment)
- a future outcome using a fitted model (e-stat11-ci-pi)
- the benefits of randomization for data collection (e-stat12-randomization)
Apply
- the principles of curiosity and skepticism to perform EDA (e-stat00-eda-basics)
Quantify
- an approximate probability with Monte Carlo (e-stat02-probability)
- the degree of uncertainty in an estimate (e-stat06-clt, e-stat09-bootstrap)
- the degree of uncertainty in a prediction (e-stat11-ci-pi)
Fit
- a distribution (e-stat08-fit-dist)
Question
- whether observed variability is real or induced (e-model00-source)
- whether a given degree of certainty is sufficient to make a decision (e-stat10-hyp-intro)
- whether a model is overfit to a given dataset (e-model02-train-validate)
- whether a model interpretation is real or spurious (e-model03-interp-warnings)

Fit
- a linear model (e-model01-intro)
- a logistic regression classifier (e-model04-logistic)
Assess
- whether a model is flexible enough to fit a given dataset (e-model02-train-validate)
- a linear model for interpretability (e-model03-interp-warnings)
- the accuracy of a classifier (e-model05-roc)
Select
- features based on an appropriate error estimate (e-stat02-model-train-validate)
- hyperparameters based on an appropriate error estimate (e-stat02-model-train-validate)

write easily-readable code by following a styleguide (e-comm00-style)
use active and constructive responding to encourage cordial discussion (e-comm01-responding)
tell a story using the ABT framework (e-comm02-story-basics)