EESW2019_tutorial

Materials for the short course on Statistical Data Cleaning for Business Statistics at the European Establishment Statistics Workshop 2019

---

Contents

Slot 1

Topic	time (m)
Introduction	20
Reading dirty data	30
Approximate matching	50
Data validation	50

Slot 2

Topic	time (m)
Error localization	20
Imputation	50
Adjusting	20
Monitoring	30
Wrap-up	10

Course form

The course form is highly hands-on. Each topic starts with an approximately 10-15 minute session where you run and adapt some R code. Next, I will provide background and details on what you just did. After that there is a more in-depth assignment. Depending on time and topic we will discuss the topic more in-depth after that.

Prerequisites

Bring a laptop

Participants are expected to have a basic knowledge of R/RStudio, explicitly:

• Work with the R command line and R scripts
• Read/write CSV data
• Some basic data manipulations and plots
• I highly recommend working with RStudio projects.

Software needed for the course

1. R See https://r-project.org
2. (Recommended) Rstudio

Execute the following R code to install the necessary packages.

install.packages(c(
        "validate"
      , "errorlocate"
      , "simputation"
      , "rspa"
      , "daff"
      , "jsonlite"
      , "XML"
      , "readr"
      , "stringr"
      , "lumberjack")
  , dependencies=TRUE)

---

License

This work is licensed under a Creative Commons Attribution 4.0 International License.