Analytics · Beginner
Python Data Analysis
Analyse real, messy operational data with pandas and NumPy: load it, clean it, group and join it, work with dates, and turn it into saved charts and a written finding, hands-on in Python 3.12.
About this course
Most data does not arrive clean. It arrives as a CSV export with duplicate rows, blank cells, three spellings of the same word and a few dates that cannot be real. This course teaches you to take data like that and answer questions with it, using **pandas 3.0** and **NumPy 2** on Python 3.12, the standard toolkit for data analysis in Python. You work in a terminal against `helpdesk`, a synthetic dataset generated for this course: about 5,000 support tickets from a fictional managed-service provider, plus the customers, agents and service plans they relate to. The tickets carry the flaws real exports have — 30 exact duplicate rows, missing customer ids, `N/A` and `-` placeholders, mixed-case categories, free-text `Yes/No/Y/true` booleans and a dozen tickets closed before they were opened — so you learn to find and fix problems, not just to call functions on tidy data. Across ten lessons you move from loading files and understanding dtypes, through selecting and filtering rows, cleaning, deriving new columns, grouping and pivoting, joining tables, and working with time series (monthly resampling and rolling averages), to summarising results and saving Matplotlib charts to PNG files with no screen. The last lesson is about being correct: the pandas mistakes that quietly give wrong answers — chained assignment, dtype surprises and silent NaN — and how to catch them with tests. The final project is a service-desk review: you answer a set of business questions end to end from the raw CSV, produce a clean dataset, saved charts and a one-page findings note, checked by a rubric and automated tests. Every command in the course was executed on pandas 3.0.5 and the output you see is the output that was observed.
- Content time
- 10 h 35 min
- Lessons
- 10
- Certificate
- Yes
- on completion
Lesson 1 is free. Enroll in a career path to access its full courses.
Lesson 1 is a free preview — read it without an account.
Outline
Lessons
Lesson 1: The analysis workflow, and setting up pandasFree preview
Set up pandas, NumPy and Matplotlib in a virtual environment and take a first, honest look at a messy dataset.
55 minLesson 2: Loading data, and the dtypes you get
Load CSV and JSON into pandas, read the dtypes you get, and control the pitfalls that turn ids into floats and placeholders into NaN.
1 hLesson 3: Selecting, filtering and indexing
Pick rows and columns with loc, iloc and boolean masks, and know which one to reach for.
55 minLesson 4: Cleaning data
Remove duplicates, handle missing values, fix types and standardise text and dates, in one reusable cleaning function.
1 h 20 minLesson 5: Transforming data, new columns from old
Derive columns with vectorised expressions, map and cut, and use apply only where it is genuinely needed.
1 hLesson 6: Grouping and aggregation
Answer "by category, by priority, by month" questions with groupby, named aggregation and pivot_table.
1 h 5 minLesson 7: Joining and merging datasets
Combine the tickets with the customer, agent and plan tables using the right kind of merge.
1 h 5 minLesson 8: Time series basics, resampling and rolling
Put a datetime on the index, resample to a monthly frequency, and smooth with a rolling average.
1 h 5 minLesson 9: Summarising and communicating results
Use describe and value_counts, export to CSV, and save Matplotlib charts to PNG for a report someone can act on.
1 hLesson 10: Reproducible, correct analysis
Avoid the pandas mistakes that give silently wrong answers - chained assignment, dtype surprises and silent NaN - and lock your results with tests.
1 h 10 min
Where it leads