18.3 EDA and Visualisation

Q11. What is EDA?

Exploratory Data Analysis — structured investigation of a dataset: shape, types, nulls, distributions, relationships, and written findings before modelling or reporting.

Q12. Correlation vs causation?

Correlation measures association. It does not prove that one column causes another. Festival demand can move both GMV and delivery minutes together.

Q13. Which chart for comparing cities?

A sorted bar chart (often horizontal). Titles should state the insight.

Q14. matplotlib vs seaborn?

matplotlib gives full control of figures/axes. seaborn speeds up statistical and categorical plots from tidy DataFrames and still draws on matplotlib.

Q15. How do you present to a manager?

Headline insight first, two supporting charts, one data-quality caveat, one clear ask — not a tour of every cell in the notebook.