Course · Beginner · 12 hours

Python for Data Analysis

Pandas, exploratory analysis and communicating findings that change decisions.

Overview

A practical Python course focused on the analysis loop: load, inspect, clean, aggregate, visualise, recommend — with portfolio-ready output at every step.

Learning objectives

  • Work fluently with DataFrames and typed columns
  • Run a disciplined exploratory data analysis
  • Aggregate and reshape data for reporting
  • Turn analysis into a recommendation stakeholders act on

Prerequisites

  • Basic Python syntax helps but is not required
  • Curiosity about business questions

Curriculum

Module 1 · Beginner

DataFrames and exploration

Load, inspect and reshape data with confidence.

Inspect data quality firstSelect with loc/ilocAggregate with groupby
Module 2 · Beginner

Communicating findings

Charts and narratives that drive decisions.

Choose the right encodingWrite sentence titlesStructure a recommendation

Capstone project

A subscription business is losing customers in month two and does not know which segment or why.

Customer retention analysis and recommendation

Synthetic subscription dataset — signups, sessions, cancellations (≈120k rows).

Requirements

  • Cohort retention analysis
  • Segment comparison with statistical sanity checks
  • Three prioritised recommendations
  • Reproducible notebook

Architecture

Ingest → Clean & type → Cohort build → Analysis → Narrative deck

Deliverables

  • Notebook
  • GitHub README
  • Executive one-pager
  • Recorded walkthrough

Resume bullet points

  • “Identified the month-two churn cliff in a 120k-user subscription dataset and proposed three prioritised interventions”
  • “Built a reproducible cohort retention notebook adopted as the team's standard analysis template”

Interview questions

loc vs iloc?

loc selects by label and is inclusive on slices; iloc selects by integer position and is exclusive on the end.

How do you approach a dataset you have never seen?

Shape, dtypes, nulls, duplicates, distributions, then targeted questions — document issues before analysing.

How do you present an unwelcome finding?

Lead with the finding, show the evidence, state the business impact, then offer options rather than blame.

Certificate

DCS Certificate — Python for Data Analysis

  • Complete all lessons
  • Score 80%+ across module quizzes
  • Submit the capstone notebook
  • Pass the final assessment