Cleaning Data in R
| Course Details | ||||||||
|---|---|---|---|---|---|---|---|---|
|
||||||||
| Languages | ||||||||
|
||||||||
| Outcomes | ||||||||
|
||||||||
| Price & Terms | ||||||||
|
||||||||
| Technical | ||||||||
|
||||||||
| Specification | ||||||||
|
||||||||
Learn to fix data types, duplicates, messy categories, missing values and typo-ridden records in R so your analysis starts from clean data.
Dirty data in R quietly skews results: values stored as the wrong type, dates that sit in the future, rows counted twice, and category labels with stray spaces or mixed capitalization. The lessons tackle these first, then move on to harder cases such as weights and currencies recorded in mixed units, totals that do not add up, and gaps in the data. The closing work merges two datasets whose values carry typos and different spellings by measuring how similar their strings are.
What You Will Achieve
- Convert columns to the correct type, trim strings, and drop values that fall outside a valid range
- Remove or aggregate duplicated rows so no record is counted twice in an analysis
- Tidy inconsistent category labels and text fields, check totals and ages, and treat missing values
- Join two datasets that contain typos by scoring how similar pairs of records are
Program Overview
Data Types, Duplicates and Categorical Data
Starts with type, range and uniqueness checks, then turns to category labels and free text that need consistent formatting.
- Data type conversion and string trimming
- Range and uniqueness constraints
- Membership checks and collapsing categories
- Detecting and replacing inconsistent text, including phone numbers
Uniformity, Validation and Missing Data
Covers values recorded in different units or formats, checks that related fields agree, and the ways data can be incomplete.
- Unit, date and currency uniformity
- Cross field validation of totals and ages
- Types of missingness
- Visualizing and treating missing data
Record Linkage with String Distance
Measures how far apart strings are and uses those scores to merge two restaurant review datasets into one clean dataset.
- Calculating string distance to fix typos
- Generating and blocking record pairs
- Scoring and linking pairs
- When to link records and when to join them
Certificate Awarded (Paid)
Job Outlook
Data cleaning skills apply wherever raw records feed reports or decisions, including data quality and data management work and the analysis done by marketing, finance and HR teams.
Good for: R users who keep running into messy columns, duplicate rows and mismatched names and want a repeatable set of checks before analysis.
Keep in mind: The exercises use prepared practice data, such as ride records and restaurant reviews, so applying each check to your own messy files is practice you do afterwards.
Bottom line: A structured run through R data checks that ends with matching records by string similarity, useful once dirty data starts skewing your results.
Frequently Asked Questions
What is record linkage and when is it used?
Record linkage matches records that describe the same thing but are written differently, for example a city given as an abbreviation in one file and in full in another. It is used to merge datasets that share no common identifier. The final chapter does this by calculating string similarity and scoring candidate pairs.
What does cross field validation check?
It checks that related fields in a record agree with each other rather than looking at one column alone. The lessons apply it to totals, confirming that values add up correctly, and to recorded ages, so inconsistent entries can be found before they affect results.
How does the course handle missing values?
The lessons first sort missing values by the type of missingness, then show how to visualize where the gaps fall in a dataset. They finish with ways to treat the missing entries so they do not distort the analysis that follows.
How much hands-on practice is there?
A short exercise follows each new concept, so the techniques are practiced as they are introduced. The last chapter brings them together in a larger task: joining two restaurant review datasets with spelling differences into a single clean dataset.
User Reviews
Write your review
You must be logged in to post a review.
Reviews can only be posted by registered learners. Log in or create a free account, then you will return here to post.
There are no reviews yet.