Learn Python for Data Science from Scratch -with 10 Projects

Explore Python data types and data structures, focusing on strings, integers, floats, booleans, and structures like lists, tuples, dictionaries, and sets, with indexing and slicing techniques.

Explore Python numeric data types - integers, floats, and complex numbers - through hands-on type checks, conversions between int, float, and complex, and practical examples in a Jupyter Notebook.

Explore Python lists as an ordered, versatile data structure that holds data types, including nested lists, and learn indexing, slicing, and operations like append, insert, extend, remove, sort, and reverse.

Learn Python tuples, an immutable, list-like collection; master indexing, slicing, and converting to a list to add items or joining tuples for flexible data structures.

Learn Python dictionaries, an ordered collection of key value pairs using curly braces; access data with get, keys, values, and items, and add, update, or remove entries like birth year.

Create an initial tuple of items for project 2 and display its contents to the user. Convert the tuple to a list to add items, following the hint provided.

Convert the names tuple to a list in a Jupyter Notebook, print the list, append Mary, then convert back to a tuple to update names to Tom, Jerry, and Mary.

Explore Python control flow by using conditional statements and loops to guide program execution from top to bottom, with real-life examples of late penalties and algorithmic solutions.

Explore Python conditional statements, including if, else, and elif, with practical examples using age and temperature. Apply logical operators and multi-condition logic using and, or, not.

Discover how loops enable repeating tasks, using for loops to iterate over lists and strings, while loops handle conditions, and the range function generates number sequences.

Explains how while loops execute code while a condition is true, detailing initialization, condition, and update steps with practical messaging examples.

Write a Python program that prints the first 20 numbers using conditionals and loops, replacing multiples of three with fees, multiples of five with balls, and both with fizzbuzz.

Explore Python basics for data science by building a fizzbuzz style sequence in a Jupyter notebook, using range, for loops, and modulus to handle divisibility rules.

Define and call functions with various parameters and return values in Python. Understand local and global scope, import built-in and third-party modules, and organize code into modular functions for readability.

Explore lambda functions in Python by creating small anonymous functions with the lambda keyword, supplying arguments and a single expression, and compare them to regular def functions.

Understand modules as reusable Python libraries—files with functions, variables, and classes you can import from math or pandas, alias for convenience, and even create your own arithmetic module.

Implement a function called calculate grid that maps a score to a letter grade using if-else logic: 90+ A, 80-89 B, 70-79 C, 60-69 D, below 60 F.

Create and test a calculate_grid function that assigns grades using if, elif, and else based on score, and preview upcoming data libraries like numpy, pandas, and matplotlib.

Learn to use NumPy, the numerical Python library, to create, reshape, index, and slice large multi-dimensional arrays with fast operations for data science and scientific computing.

Install numpy, import it as np, and create arrays with np.array to explore zero-dimensional, one-dimensional, two-dimensional, and three-dimensional arrays and their ndim properties.

Explore indexing and slicing NumPy arrays to access specific elements and ranges in one dimensional and two dimensional data structures, with practical examples and zero-based indexing.

Learn how to copy and view numpy arrays to manage data efficiently, using array.copy for independent copies and array.view for shared data where changes to a view affect the original.

Learn how shape defines elements per dimension in numpy arrays and reshape from 1d to 2d or 3d, including the outermost dimension and examples like two-by-five.

Explore basic numpy array operations, including arange for generating ranges, and element-wise math like addition, subtraction, multiplication, and division. Learn to apply square and square root for data analytics with numpy.

Explore data analytics with NumPy: perform array operations, summations, min/max, and axis-based computations on 1d and 2d arrays, including mean across rows and columns.

Import numpy, then generate a numpy array filled with random numbers by prompting the user for rows, columns, min and max, and print the array to the console.

Create a Python program that prompts for rows, columns, min and max values, builds a NumPy array of that shape with random integers in the range, and prints it.

Explore pandas, the open source Python data analysis and manipulation library, and learn to create series and data frames, clean data, and perform grouping and aggregation.

Learn how to read in files with pandas to ingest diverse data sources, producing data frames from CSV, JSON, Excel, and more using read functions and file paths.

Explore a dataframe using pandas in a Jupyter notebook by loading data with read_csv, then view head and tail, inspect shape, and use info and describe for basic statistics.

Learn to set and change a pandas data frame index with country, use loc and iloc for label and integer based searches, and reset the index to its original form.

Discover how to group by base flavor in Pandas and apply aggregate functions such as mean, min, max, and count across flavor, texture, and total rating.

Explore how to merge, join, and concatenate data frames in pandas using common columns or indices, with inner, outer, left, and right joins.

Learn data cleaning in pandas by handling nulls with dropna and fillna, replacing missing values with mean or median, and converting dates to datetime to fix bad data.

Address missing values and correct wrong imputations in pandas using loc to fix specific cells. Clean outliers by updating or dropping rows, and remove duplicates with drop_duplicates.

Learn to visualize data in pandas using the plot function, creating line, bar, histogram, and pie charts while cleaning data, setting a date time index, and exploring subplots.

Analyze a world population dataset with pandas, using groupby to compute average populations by continent for 1970–2022, sort by 2022, transpose, and plot the results as a line chart.

Load a world population dataset in Jupyter, group by continent with pandas, compute mean populations from 1970 to 2022, and plot the results with a descending 2022 value.

Explore Matplotlib, a Python library for creating visualizations in data science, and learn to build basic plots, customize visuals, arrange multiple subplots, and save figures in various formats.

Explore basic matplotlib plotting with pandas in a Jupyter notebook, creating line, bar, and scatter visuals from a calories dataset and its data frames, including cleaning and labeling.

Visualize London daily weather statistics using matplotlib, resample daily data to yearly with pandas, and create exploratory plots of time series weather insights.

Solve project 7 by loading London weather data with pandas, cleaning and converting dates, resampling to yearly means, and visualizing annual trends for sunshine, temperature, and precipitation.

Introduce machine learning fundamentals and how models learn from data to make predictions, using scikit-learn to load, explore, and pre-process data—including missing values and feature scaling—for real-world applications.

Master foundational machine learning techniques, including deciding if ML is the right approach, data collection and preparation, train-test splits, model training, evaluation, hyperparameter tuning, and real-world deployment.

Explore scikit-learn, a core machine learning library for building, training, and evaluating models in data science, covering classification, regression, and clustering; install, import, and use it in Jupyter notebooks.

Master linear, polynomial, and logistic regression with scikit-learn, apply feature selection techniques, and evaluate models using mean squared error and r-squared for house and stock price predictions.

Build and evaluate your first linear regression model to predict future outcomes from historical data, understand features and targets, and practice train–test splits using Python and basic libraries.

Build a linear regression model in python to predict sales from radio advertising, using tv, radio, and newspaper spend, and split the data with train-test split while visualizing with matplotlib.

Build and evaluate a linear regression model using radio ad spend to predict sales, training on a split dataset and assessing via mean absolute error and mean squared error.

Save and load linear regression models with joblib and pickle files, reuse them in other notebooks, and evaluate performance using mean absolute error and mean squared error.

Explore predicting sales from TV advertising budgets using linear regression and polynomial regression on the course dataset, and compare the fits in project 8.

Explore how classification models sort data into two groups using shape and color, evaluate with accuracy, precision, recall, and F1, and tune hyperparameters to optimize supervised learning performance.

Learn to build a wine quality classification model in a notebook, using fixed acidity and pH, and compare logistic regression, random forest, and SVC with accuracy, recall, and F1.

Learn how to clean wine quality data, drop unnecessary columns, create a binary high/low quality label, and split features x and target y for your first classification model.

Identify data imbalance in wine quality data and apply smote oversampling to balance classes, then split balanced data into train and test sets for model training.

Build and evaluate a resampled balanced binary classifier using a random forest in Python, training on the training data, predicting test results, and interpreting accuracy, precision, recall, and F1.

Experiment with three classification models on the White Wine Quality dataset, compare their metrics, and determine which model performs best.

Predict wine quality using the white wine dataset by preprocessing a semicolon-delimited csv, encoding the target, balancing data, and comparing logistic regression, random forest, and SVC.