liveBook · Manning

Realistic data

As explained in Everyday Data Visualization

Realistic data has structures, values, and distributions that plausibly resemble those of the intended real-world data.

Realistic data resembles the intended real-world data in its columns, values, and distributions. It helps prototypes behave like finished products before live data arrives. section 9.1.2

Shape test data around the conditions the final product will face:

  • Choose believable columns, including dates, times, names, and addresses when relevant. section 9.1.2
  • Constrain values to the expected setting, such as one country or future dates. section 9.1.2
  • Match distributions, such as subscription prices or house prices clustered around a regional average. section 9.1.2
  • Use a random generator or an LLM, then fine-tune the results instead of accepting convenient nonsense. section 9.1.2

A survey prototype

A generated survey dataset can preserve respondent rows, question text, and scaled responses. Two respondents answering three questions provide a small but structurally believable fixture for a prototype. section 9.1.2

Table 9.1 A simple example of a survey with three questions and two respondents, generated by ChatGPT (view table figure)

Respondent

Question text

Question response

1

"On a scale from 1 to 5, how likely are you to recommend our app to a friend?"

4

1

"Please rate the app on a scale from 1 to 10."

7

1

"On a scale from 1 to 10, how much did you enjoy your breakfast this morning?"

8

2

"On a scale from 1 to 5, how likely are you to recommend our app to a friend?"

3

2

"Please rate the app on a scale from 1 to 10."

6

2

"On a scale from 1 to 10, how much did you enjoy your breakfast this morning?"

6

Browse a generated survey table showing question structure and responses for two respondents. section 9.1.2

How Everyday Data Visualization covers it

How deep it goes
Defines realistic columns, values, and distributions, then warns that structural mismatch can force a rebuild.
Examples
Shows survey responses, subscription prices, regional house prices, dates, times, names, and addresses.
What you can do after
Apply it to generate test data that can transition more smoothly into real data sources.
How it connects
Connects test data with prototypes, random data generators, large-language models, data sources, and visualization.
Cover of Everyday Data Visualization
liveBookmanning.com
Read Everyday Data Visualization on liveBook Get the book at manning.com

This explainer was generated by Manning's knowledge graph from the text of Everyday Data Visualization. The section links point to the parts of the book it draws on.