earndws

Workshop Materials for the EARN Data Bootcamp

earndws is an R package that bundles scripts, data, and dependencies for the Economic Analysis and Research Network (EARN) Conference data bootcamp workshops. Its single setup function creates a ready-to-run working directory so participants can start coding without any manual file downloads or directory setup.

---

Installation

Install the package from its source directory (or from GitHub if hosted there):

# Using remotes/pak if hosted on GitHub
remotes::install_github("Economic/earndws")

Required packages

The following packages are installed automatically as dependencies:

Package	Purpose
tidyverse	Data wrangling and visualization
data.table	Variable renaming via setnames()
lubridate	Date parsing and manipulation
openxlsx2	Excel workbook creation
haven	Factor handling from survey data
maps	U.S. geographic map data

---

Quick Start

Load the package and call setup_workshop() with the name of your workshop:

library(earndws)

# Set up in the current working directory
setup_workshop("foraging-for-data")

# Or set up in a specific folder
setup_workshop("intro-to-r", path = "~/earn-workshop")

That's it. The function prints a confirmation message and your working directory is ready:

Workshop ready in '/home/you/earn-workshop'
  code/   - workshop scripts
  input/  - data files
  output/ - your results go here

Open any script in code/ and run it — all data files it needs are already in input/.

---

setup_workshop()

setup_workshop(workshop, path = ".")

Arguments

Argument	Type	Description
workshop	character	One of "foraging-for-data" or "intro-to-r"
path	character	Target directory. Defaults to current working directory (.)

What it does

1. Creates three subdirectories inside path: code/, input/, and output/
2. Copies the workshop R scripts into code/
3. Writes the relevant bundled datasets as CSVs into input/ — no internet connection needed

Returns NULL invisibly (called for its side effects).

---

Workshops

"foraging-for-data"

An applied data wrangling workshop using weekly unemployment insurance (UI) claims data from the U.S. Department of Labor's ETA 539 report.

What you'll work with

• Raw ETA 539 claims data with DOL short-code column names
• A data dictionary that maps those codes to human-readable titles
• State-level initial and continued UI claims (non-seasonally adjusted), filtered to weeks from 1987 onward

Scripts

Script	Description
state_ui.R	Full participant script: loads data, renames variables using the DOL data dictionary, calculates NSA initial and continued claims, pivots to wide format (states as columns), adds a US Total column to continued claims, and exports a formatted two-sheet Excel workbook to output/state_ui.xlsx
live-demo-practice.R	Streamlined live-demo reference with the same core workflow but no output formatting — intended as an instructor companion during the session

Bundled data written to input/

File	Description
ar539.csv	Weekly state-level UI initial and continued claims from the DOL ETA 539 report. One row per state per week; column names use DOL short codes (e.g. c1, c3, rptdate)
eta539_var_names.csv	Data dictionary with two columns — dol_code and dol_title — used to rename ar539 columns to readable labels

---

"intro-to-r"

A fundamentals workshop teaching R through hands-on analysis of national and state unemployment rates from CPS Outgoing Rotation Group (ORG) data.

Note: This workshop requires the epiextractr package. See Optional: epiextractr above.

What you'll learn

• Basic R syntax and arithmetic
• Vectors, functions, and variable assignment
• Data frames: subsetting with $ and [,], filtering with subset()
• The pipe operator (|>) and readable data workflows
• Core tidyverse verbs: filter(), mutate(), summarise()
• Weighted statistics: weighted.mean() with survey weights
• Conditional logic with if_else()
• Working with factors and string operations
• Data visualization with ggplot2: bar charts and choropleth maps

Scripts

Script	Description
intro_to_r.R	Comprehensive participant tutorial walking through all concepts from basic arithmetic to geographic mapping, with explanatory comments throughout
live_code.R	Condensed live-coding reference with key examples for each concept — intended as an instructor companion during the session

Bundled data written to input/

File	Description
unemp_rates.csv	National unemployment rates derived from CPS ORG extracts. Columns: statefips, unemp_rate, usps
unemp_rates_state.csv	State-level unemployment rates from CPS ORG extracts. Same structure as above

---

Directory Structure After Setup

your-path/
├── code/
│   ├── state_ui.R                  # (foraging-for-data)
│   └── live-demo-practice.R
│   # or
│   ├── intro_to_r.R                # (intro-to-r)
│   └── live_code.R
├── input/
│   ├── ar539.csv                   # (foraging-for-data)
│   └── eta539_var_names.csv
│   # or
│   ├── unemp_rates.csv             # (intro-to-r)
│   └── unemp_rates_state.csv
└── output/
    └── (your results go here)

---

Bundled Datasets

The package also exposes its datasets directly as R objects after library(earndws):

# ETA 539 UI claims data
?ar539

# DOL variable name dictionary
?eta539_var_names

# National CPS unemployment rates
?unemp_rates

# State-level CPS unemployment rates
?unemp_rates_state

---

Author

Jori Kandra — jkandra@epi.org
Economic Policy Institute