Data Science Fundamentals
The data science process in plain Python: reading, cleaning and reshaping data, summaries, groups and joins, relationships, time series, prediction and evaluation.
- Level
- Intermediate
- Time
- About 28 hours
- Modules
- 5
- Lessons
- 15
- Exercises
- 45
Data science turns raw records into answers people can act on. This course teaches the whole process, from a vague question and a messy file to a cleaned table, honest summaries, a first predictive model and a clear report.
Everything is written in plain Python, so you see exactly what tools such as pandas and scikit-learn do for you. Every example runs step by step in your browser, and every exercise is checked against hidden tests.
What you will learn
- Turn a question into an analysis plan and read CSV and JSON data
- Clean, reshape and validate messy data and log every decision
- Summarise, group and join tables to answer questions
- Explore distributions, relationships and time series without being misled
- Build and evaluate simple predictive models against a baseline
- Run a reproducible, honest analysis from raw data to report
Skills you will gain
- Data cleaning
- Tidy data and reshaping
- Descriptive statistics
- Grouping and joining
- Exploratory analysis
- Model evaluation
Before you start
- The Python course, or comfort with lists, dictionaries, loops and functions
- Basic statistics helps: the mean, the median and percentages
- Nothing to install: all code runs in your browser
Where it is used
- Business and product analytics
- Reports and dashboards
- Research data analysis
- A foundation for pandas and machine learning
Modules
Module 1: Getting started with data
What data science is, how tables, types and variables work, and how to read CSV and JSON data with plain Python.
- 1
What data science is Intermediate
The data science process from a question to an answer: asking a precise question, collecting and cleaning data, exploring it, modelling it and communicating what was found, with a first analysis in plain Python.
- 2
Data types and tables Intermediate
The kinds of variables a dataset holds (numeric, categorical, ordinal, yes/no, text and dates), how a table is organised in rows and columns, and how to inspect a table's shape, types and missing values.
- 3
Reading CSV and JSON data Intermediate
Loading the two most common data formats with the standard library: CSV with csv.reader and csv.DictReader, including quoted fields, JSON with json.loads and json.dumps, and converting text values to the right types.
Module 2: Cleaning and shaping data
Handling missing values, duplicates and inconsistent entries, reshaping between wide and long tables, and working with dates and text.
- 1
Cleaning data Intermediate
Finding and fixing the problems real data has: missing values, inconsistent text, impossible values and duplicates, and the choices between dropping, flagging and filling, with mean, median or a default.
- 2
Tidy data: reshaping tables Intermediate
The tidy data principles (each variable a column, each observation a row, each value a cell), melting wide tables into long ones, pivoting long tables into wide ones, and splitting columns that hold two variables.
- 3
Dates, times and text Intermediate
Parsing dates and times with datetime and strptime, extracting parts such as the weekday, measuring differences with timedelta, and normalising free text so that the same answer written differently is counted once.
Module 3: Summarising and combining
Describing columns with centre and spread, grouping and aggregating, and joining tables on keys.
- 1
Descriptive statistics Intermediate
Summarising a numeric column: the mean, median and mode for its centre; the range, variance, standard deviation, quartiles and IQR for its spread; and finding outliers with the 1.5 × IQR rule.
- 2
Grouping and aggregation Intermediate
Split, apply, combine: grouping rows by one or more keys with dictionaries, aggregating each group with counts, sums, means, minimums and maximums, sorting the results, and building cross tabulations.
- 3
Joining tables Intermediate
Combining tables through a shared key: building a dictionary index, inner joins that keep matches, left joins that keep every row of one table, many-to-one relationships, and finding rows that match nothing.
Module 4: Exploring relationships and time
Exploratory data analysis with profiles and histograms, relationships between variables and their traps, and time series basics.
- 1
Exploratory data analysis Intermediate
Getting to know a dataset before modelling it: profiling every column, counting missing values, drawing text histograms with bins, and writing down questions and surprises as you go.
- 2
Relationships between variables Intermediate
How two columns move together: scatter thinking, Pearson's correlation for every pair of columns, confounders, spurious correlations and Simpson's paradox, where a trend reverses when groups are combined.
- 3
Time series basics Intermediate
Data measured over time: sorting by date, totals per month, percentage change, moving averages that smooth out noise, seasonality, and flagging unusual days with a rolling mean and standard deviation.
Module 5: Prediction and practice
Training and test sets, baselines and simple models, evaluating classifiers, and running a reproducible, honest analysis.
- 1
Prediction basics Advanced
The core ideas of predictive modelling in plain Python: features and targets, training and test sets, a baseline to beat, a least-squares line, nearest neighbours, and error measures such as MAE and RMSE.
- 2
Evaluating classifiers Advanced
Judging yes/no predictions properly: the confusion matrix, accuracy and why it misleads on rare classes, precision, recall, the F1 score, and choosing a decision threshold.
- 3
The data science workflow Advanced
Putting the steps together into a reproducible analysis: a pipeline from raw data to report, train and test discipline, data leakage, documenting cleaning decisions, ethics and privacy, and communicating results honestly.
Projects
- A sales data analysis A shop exports a year of messy sales records. Clean them, answer which products, regions and months drive sales, and write a short report with the cleaning log.
- A customer churn predictor Predict which customers of a subscription service will leave, using a few features, and decide whether the model is good enough to act on.
What to learn next
- Machine Learning Machine learning from scratch in plain Python: regression, gradient descent, regularization, k-NN, logistic regression, metrics, trees, ensembles and k-means.
- Statistics & Probability Chance and data in Python: counting, Bayes, simulation, distributions, sampling, confidence intervals, hypothesis tests, regression and chi-square tests.
- Database Management Systems Tables and SQL: queries, joins, grouping, keys, subqueries, ER design, normalization, relational algebra, indexes and transactions, on a real database in your browser.