Data Science

Data Science — Zero to Advanced

Python, NumPy, pandas, statistics, and predictive modeling — 53 lessons, one live in-browser dataset, no prerequisites

Data science cheat sheet and project ideas →

Self-paced July 2026

🎓Complete beginners — zero Python or stats knowledge needed

📊Analysts moving from Excel/SQL into pandas and statistics

💼Anyone preparing for data science interview questions

🔄Engineers who want to add data science to their toolkit

Lessons

Sections

213+

Topics covered

14h

Total content

100%

Free forever

This track teaches data science from first principles. Every concept is introduced with a real dataset, practiced immediately in a live Python notebook running in your browser, and connected to a real job context. You write real pandas from Lesson 01 — no setup, no install, no account required.

What makes this different

▶

Live Python Playground

Run real pandas code in your browser — no install, no account. Uses Pyodide with the StreamPulse dataset preloaded.

Open →

◎

Try It Challenges

Every lesson ends with a practice question. Reveal the answer and explanation when you're ready.

Open →

📈

Real Charts, Live

Matplotlib and seaborn figures render right in the playground — you build the chart, not just read about it.

Open →

≡

One Dataset, 53 Lessons

StreamPulse — a fictional streaming service — is used from your first line of pandas to your final capstone project.

Open →

Your learning dataset

The StreamPulse Dataset

Every module, every notebook cell, every exercise uses StreamPulse — a fictional video streaming service with subscribers across 8 countries, a 20-title catalog, and real engagement and churn patterns.

users25 rows

user_idint64

first_nameobject

last_nameobject

emailobject

countryobject

+5 more columns

titles20 rows

title_idint64

title_nameobject

typeobject

genreobject

release_yearint64

+4 more columns

subscriptions25 rows

subscription_idint64

user_idint64

planobject

start_datedatetime64

end_datedatetime64

+3 more columns

watch_history100 rows

watch_idint64

user_idint64

title_idint64

watch_datedatetime64

minutes_watchedint64

+2 more columns

ratings40 rows

rating_idint64

user_idint64

title_idint64

ratingint64

rating_datedatetime64

+1 more columns

// Curriculum

53 Lessons. Zero to Advanced.

Follow in order. Each module builds on the last. Lesson 01 assumes you know nothing — Lesson 53 ends with a portfolio-ready capstone project and interview prep.

Section 1 — Data Science Foundations

MODULE 01✓ LIVE

What is Data Science?

The definition that actually explains it, the DS lifecycle, and why every streaming, ride-hailing, and fintech app runs on this discipline

Data Science definitionDS lifecycleWhy it mattersStreamPulse intro

⏱ 10–14 min

Beginner

Start →

MODULE 02✓ LIVE

The Data Science Workflow

From a vague business question to a shipped decision — the six stages every real project moves through

Problem framingData collectionEDAModelingCommunication

⏱ 12–16 min

Beginner

Start →

MODULE 03✓ LIVE

Data Science vs Data Engineering vs ML Engineering vs Analytics

Four job titles that get confused constantly — what each one actually owns day to day

DS vs DEDS vs ML EngDS vs AnalystCareer paths

⏱ 10–14 min

Beginner

Start →

MODULE 04COMING SOON

Setting Up — Your Browser-Based Python Environment

Meet the live Python + pandas playground running entirely in your browser, and the StreamPulse dataset you will use for every module

PyodideJupyter alternativepandasStreamPulse dataset

⏱ 8–12 min

Beginner

Soon

Section 2 — Python Foundations for Data Science

MODULE 05COMING SOON

Variables and Data Types

Numbers, strings, booleans, and None — the building blocks of every line of analysis code

Variablesint/float/strBooleansType checking

⏱ 10–14 min

Beginner

Soon

MODULE 06COMING SOON

Lists, Dictionaries, and Tuples

The containers that hold real datasets before they ever reach a DataFrame

ListsDictionariesTuplesIndexing

⏱ 12–16 min

Beginner

Soon

MODULE 07COMING SOON

Control Flow — if/else and Loops

Decisions and repetition — the logic every data cleaning script depends on

if/elif/elsefor loopswhile loopsbreak/continue

⏱ 10–14 min

Beginner

Soon

MODULE 08COMING SOON

Functions and List Comprehensions

Write reusable logic once, and the one-line pattern every data scientist uses instead of loops

def functionsArgumentsList comprehensionsLambda

⏱ 12–16 min

Beginner

Soon

MODULE 09COMING SOON

Working with Files, CSV, and JSON

Reading and writing the two formats that carry almost all real-world data

File I/OCSVJSONEncoding

⏱ 10–14 min

Beginner

Soon

Section 3 — NumPy — Numerical Computing

MODULE 10COMING SOON

NumPy Arrays — The Foundation of Everything

Why every numeric library in Python is built on top of the ndarray

ndarrayCreating arraysShape & dtypeWhy not lists

⏱ 12–16 min

Beginner

Soon

MODULE 11COMING SOON

Indexing, Slicing, and Boolean Masks

Select exactly the elements you need — the pattern that also powers pandas filtering

IndexingSlicingBoolean masksFancy indexing

⏱ 12–16 min

Beginner

Soon

MODULE 12COMING SOON

Vectorization and Broadcasting

Why NumPy is 100x faster than a Python loop, and the broadcasting rules that make it work

VectorizationBroadcasting rulesPerformanceElement-wise ops

⏱ 14–18 min

Beginner

Soon

MODULE 13COMING SOON

Aggregations and Basic Statistics

mean, std, sum, min, max, and the axis parameter that trips up every beginner

mean/std/sumaxis parameterargmin/argmaxCumulative functions

⏱ 12–16 min

Beginner

Soon

Section 4 — Pandas Fundamentals

MODULE 14COMING SOON

Series and DataFrames

The two core pandas objects — and how the StreamPulse tables become live DataFrames the moment this page loads

SeriesDataFrameIndexdtypes

⏱ 12–16 min

Beginner

Soon

MODULE 15COMING SOON

Reading and Inspecting Data

head, tail, info, describe, shape — the five commands you run on any dataset in the first ten seconds

read_csv.head()/.info().describe()First look at data

⏱ 10–14 min

Beginner

Soon

MODULE 16COMING SOON

Selecting and Filtering Rows

loc, iloc, and boolean filtering — getting exactly the rows you need out of any DataFrame

.loc/.ilocBoolean filteringMultiple conditionsquery()

⏱ 14–18 min

Beginner

Soon

MODULE 17COMING SOON

Sorting and Ranking

sort_values, sort_index, rank — controlling the order your analysis surfaces results in

sort_valuessort_indexrank()nlargest/nsmallest

⏱ 8–12 min

Beginner

Soon

MODULE 18COMING SOON

Adding and Modifying Columns

Derived columns, renaming, dropping — shaping a DataFrame into the form your analysis needs

Column assignmentrename()drop()assign()

⏱ 10–14 min

Beginner

Soon

Section 5 — Data Cleaning

MODULE 19COMING SOON

Handling Missing Values

isna, dropna, fillna — the decision tree every data scientist runs before any analysis

NaNisna()/notna()dropna()fillna()

⏱ 14–18 min

Intermediate

Soon

MODULE 20COMING SOON

Duplicates and Data Type Conversion

Finding and removing duplicate rows, and casting columns to the type they should have been all along

duplicated()drop_duplicates()astype()to_datetime()

⏱ 12–16 min

Intermediate

Soon

MODULE 21COMING SOON

String Cleaning and Regular Expressions

Whitespace, casing, and pattern extraction — the .str accessor and regex basics for messy text columns

.str accessorstrip/lower/replaceRegex basicsextract()

⏱ 14–20 min

Intermediate

Soon

MODULE 22COMING SOON

Outlier Detection

IQR, z-scores, and visual methods for finding the data points that will break your model

IQR methodZ-scoreBoxplotsWhen to remove vs keep

⏱ 14–18 min

Intermediate

Soon

MODULE 23COMING SOON

Data Validation Checks

Building a checklist that catches bad data before it reaches a stakeholder or a model

Schema checksRange checksReferential checksAssertions

⏱ 10–14 min

Intermediate

Soon

Section 6 — Data Wrangling

MODULE 24COMING SOON

GroupBy Fundamentals

The single most powerful pandas operation — split, apply, combine, explained visually

groupby()Split-apply-combineagg()Single-column grouping

⏱ 16–22 min

Intermediate

Soon

MODULE 25COMING SOON

Multi-Column GroupBy and Aggregation

Grouping by more than one column and applying different aggregations to different columns at once

Multi-key groupbyNamed aggregationtransform()filter()

⏱ 14–18 min

Intermediate

Soon

MODULE 26COMING SOON

Merging and Joining DataFrames

Combining users, watch_history, and titles into one analysis-ready table — inner, left, right, outer

merge()Inner/left/right/outerjoin()Merge keys

⏱ 16–22 min

Intermediate

Soon

MODULE 27COMING SOON

Concatenating and Reshaping

Stacking DataFrames together and reshaping between wide and long formats

concat()Wide vs longstack/unstackAxis parameter

⏱ 12–16 min

Intermediate

Soon

MODULE 28COMING SOON

Pivot Tables and melt()

Excel-style pivot tables in pandas, and the reverse operation that tidies messy spreadsheets

pivot_table()melt()Tidy dataCross-tabulation

⏱ 14–18 min

Intermediate

Soon

MODULE 29COMING SOON

apply, map, and Lambda Functions

Running custom logic across rows and columns when built-in pandas methods are not enough

apply()map()applymap()Lambda functions

⏱ 12–16 min

Intermediate

Soon

Section 7 — Exploratory Data Analysis

MODULE 30COMING SOON

Descriptive Statistics

Mean, median, mode, variance, standard deviation — and which one lies to you when data is skewed

Mean/median/modeVariance & std devSkewnessdescribe()

⏱ 14–18 min

Intermediate

Soon

MODULE 31COMING SOON

Distributions and Histograms

Seeing the shape of your data before you trust any summary statistic about it

HistogramsBinsSkew & kurtosisNormal vs skewed

⏱ 12–16 min

Intermediate

Soon

MODULE 32COMING SOON

Correlation Analysis

Pearson correlation, correlation matrices, and why correlation is not causation — with a StreamPulse example

corr()Correlation matrixHeatmapsCorrelation vs causation

⏱ 14–18 min

Intermediate

Soon

MODULE 33COMING SOON

A Complete EDA Workflow — Case Study

Start to finish: exploring StreamPulse watch_history to find what actually drives engagement

EDA checklistCase studyHypothesis generationEngagement analysis

⏱ 18–24 min

Intermediate

Soon

MODULE 34COMING SOON

Storytelling with Data

Turning a notebook full of charts into a narrative a non-technical stakeholder will act on

Data storytellingStructuring an analysisSlide-ready chartsAvoiding chart junk

⏱ 12–16 min

Intermediate

Soon

Section 8 — Data Visualization

MODULE 35COMING SOON

Matplotlib Fundamentals

Figures, axes, and the object-oriented API that every other Python plotting library builds on

Figure & Axesplot()Labels & titlesSubplots

⏱ 14–18 min

Intermediate

Soon

MODULE 36COMING SOON

Seaborn — Statistical Plots Made Simple

Boxplots, violin plots, pairplots, and heatmaps in one line instead of twenty

Seaborn basicsBoxplot/violinplotPairplotHeatmap

⏱ 14–18 min

Intermediate

Soon

MODULE 37COMING SOON

Choosing the Right Chart

A decision framework for picking bar vs line vs scatter vs box — and the charts that mislead

Chart selectionBar vs line vs scatterMisleading chartsChart anatomy

⏱ 10–14 min

Intermediate

Soon

MODULE 38COMING SOON

Building a Dashboard View

Combining multiple charts into one cohesive summary of StreamPulse subscriber health

Subplots layoutKPI summaryDashboard designAnnotations

⏱ 14–20 min

Intermediate

Soon

Section 9 — Statistics and Probability

MODULE 39COMING SOON

Probability Basics

Events, sample spaces, conditional probability, and Bayes theorem — with StreamPulse churn examples

Sample spaceConditional probabilityBayes theoremIndependence

⏱ 16–22 min

Advanced

Soon

MODULE 40COMING SOON

Common Probability Distributions

Normal, binomial, and Poisson — the three distributions that explain most real-world data

Normal distributionBinomial distributionPoisson distributionPDF vs CDF

⏱ 16–22 min

Advanced

Soon

MODULE 41COMING SOON

The Central Limit Theorem

Why sample means are normally distributed even when the underlying data is not — and why this matters for every test that follows

CLTSampling distributionStandard errorLaw of large numbers

⏱ 14–18 min

Advanced

Soon

MODULE 42COMING SOON

Hypothesis Testing

Null and alternative hypotheses, t-tests, p-values — and what a p-value actually does not mean

Null/alternative hypothesist-testp-valuesType I & II errors

⏱ 18–24 min

Advanced

Soon

MODULE 43COMING SOON

Confidence Intervals

Quantifying uncertainty around an estimate — and why "95% confident" does not mean what most people think

Confidence intervalsMargin of errorInterpretationSample size effects

⏱ 14–18 min

Advanced

Soon

MODULE 44COMING SOON

A/B Testing Fundamentals

Designing a valid experiment to test a new StreamPulse recommendation algorithm end to end

A/B test designControl vs treatmentStatistical significanceCommon pitfalls

⏱ 18–24 min

Advanced

Soon

Section 10 — Intro to Predictive Modeling

MODULE 45COMING SOON

Train/Test Split and Overfitting

Why every model must be evaluated on data it has never seen — and what overfitting looks like in practice

train_test_splitOverfitting vs underfittingBias-variance tradeoffValidation sets

⏱ 16–22 min

Advanced

Soon

MODULE 46COMING SOON

Linear Regression from Scratch

Predicting watch time from subscriber attributes — the math, then scikit-learn

Linear regressionCoefficientsscikit-learn basicsR-squared

⏱ 18–24 min

Advanced

Soon

MODULE 47COMING SOON

Classification with Logistic Regression

Predicting subscriber churn — probability outputs, decision thresholds, and the sigmoid function

Logistic regressionSigmoid functionDecision thresholdBinary classification

⏱ 18–24 min

Advanced

Soon

MODULE 48COMING SOON

Model Evaluation Metrics

Accuracy, precision, recall, F1, ROC-AUC — and which one actually matters for a churn model

Confusion matrixPrecision/recallF1 scoreROC-AUC

⏱ 16–22 min

Advanced

Soon

MODULE 49COMING SOON

Intro to Feature Engineering

Turning raw StreamPulse columns into signals a model can actually learn from

Feature creationEncoding categoricalsScalingFeature selection basics

⏱ 16–22 min

Advanced

Soon

Section 11 — Real-World Case Studies and Career

MODULE 50COMING SOON

Churn Analysis — Full Case Study

Start to finish: why StreamPulse subscribers cancel, using everything from Modules 1–49

Case studyChurn driversEnd-to-end analysisStakeholder summary

⏱ 22–30 min

Advanced

Soon

MODULE 51COMING SOON

Cohort and Retention Analysis

The analysis every subscription business runs monthly — signup cohorts, retention curves, and churn by segment

Cohort analysisRetention curvesSignup cohortsSegment comparison

⏱ 18–24 min

Advanced

Soon

MODULE 52COMING SOON

Data Science Interview Questions

30 real interview questions with complete answers — statistics, pandas, SQL crossover, and case study rounds

Interview prepStatistics questionsPandas questionsCase study rounds

⏱ 24–30 min

Advanced

Soon

MODULE 53COMING SOON

Building Your Portfolio Project

Turning a StreamPulse-style analysis into a portfolio piece that gets you hired

Portfolio projectsGitHub presentationProject write-upsWhat recruiters look for

⏱ 14–20 min

Advanced

Soon

// Ready to start?

Start with Lesson 01. No setup required.

Every module runs entirely in your browser. Write real Python and pandas against the StreamPulse dataset from your very first lesson.

Start Lesson 01 →

Discussion

Have a better approach? Found something outdated? Share it — your knowledge helps everyone learning here.

Continue with GitHub