liveBook · Manning

Known-question analysis

As explained in Vibe Engineering

Known-question analysis is a data-analysis approach in which predefined questions determine the computations and outputs to produce.

Known-question analysis starts with predefined questions, then translates each question into explicit computations, outputs, and visualizations. This reduces ambiguity when an LLM generates analytical code. section 4.3.1

A practical specification identifies:

  • The required dataset path and filename, preventing generated notebooks from searching the wrong directory. section 4.3.1
  • The grouping, ranking, aggregations, percentages, tables, and charts needed to answer each question. section 4.3.1
  • Export requirements such as an images directory and high-resolution PNG output. section 4.3.1

Pharmaceutical market analysis

A defined analysis ranks the ten manufacturers with the most products, compares their average prices, measures discontinued products and their percentage, and summarizes count and average price by medicine type. section 4.3.1

Figure 4.3 Top 10 drug manufacturers ranked by number of products sold, with each manufacturer’s average drug price shown alongside.

Compare manufacturer product counts with average prices in a dual-axis visualization. section 4.3.1

Listing 4.2 The Jupyter analysis output
--- Medicine Type Analysis Summary ---
            Count  Average Price (INR)
type                                  
allopathy  253973           270.530844
 

Inspect the generated summary table combining medicine-type counts and average prices. section 4.3.1

How Vibe Engineering covers it

How deep it goes
Specifies computations, charts, tables, paths, and exports rather than merely naming analytical questions.
Examples
Uses pharmaceutical manufacturers, prices, discontinued products, medicine types, charts, and a summary DataFrame.
What it assumes
Assumes predefined questions and a correctly located dataset, while teaching their translation into an executable notebook prompt.
What you can do after
Build a specified Jupyter analysis that computes requested metrics, generates plots, and exports PNG artifacts.
How it connects
Connects data analysis with LLM prompting, Python, Jupyter Notebook, and data visualization.

Related concepts

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