Practical Work and Project
What this domain tests
The practical examination checks whether you can create and query a working database, run a genuine Pandas analysis on a CSV file, and present the work in a structured project report with interpretation.
SQL practical
Create the database and table with suitable data types, a primary key and NOT NULL where required, insert sample rows, then run queries that use WHERE, ORDER BY, aggregates with GROUP BY and HAVING, and at least one join. Always test the query and check the row count against the expected answer.
Pandas practical
Load the CSV with read_csv, inspect with head(), shape and info(), handle missing values, compute grouped summaries with groupby, merge a second table if relevant, and plot the result. Save the cleaned data with to_csv(..., index=False).
Project report structure
Title and objective; problem statement; data source and its limitations; tools used; methodology; annotated code; output tables and charts; findings tied to the objective; limitations; conclusion; bibliography. Each chart needs a sentence saying what it shows, not just its caption.
Worked example
A project on class performance reads marks.csv (Roll, Subject, Marks). Steps: fill missing marks with the subject mean after checking how many are missing; groupby('Subject')['Marks'].mean(); merge with a student names table on Roll; draw a bar chart of subject means with axis labels and a title; report that one subject is weakest and recommend remedial work. Reasoning: each step answers a stated objective, and the recommendation follows from the computed means.
Traps that cost marks
Leaving or hardcoding the row index in to_csv; using absolute file paths that break on the examiner's machine; code that only runs after earlier hidden cells; SQL queries with a missing GROUP BY column; screenshots with no interpretation; unsupported claims; no citation of the data source; a conclusion that ignores the findings.