Welcome to ISLP documentation!

ISLP is a Python library to accompany Introduction to Statistical Learning with applications in Python. See the statistical learning homepage for more details.

See the ISLP reference

Contents

• Install instructions
  • Mac OS X / Linux
  • Windows
• Installing ISLP
  • Frozen environment
  • Torch requirements
  • Jupyter
    • Mac OS X
    • Windows
• Google Colab
• Datasets used in ISLP
  • Auto Data
    • Notes
  • Bike sharing data
    • Source
  • Boston Data
    • Notes
  • Brain Cancer Data
    • Source
  • Caravan
    • References
  • Sales of Child Car Seats
  • U.S. News and World Report’s College Data
    • Source
  • Credit Card Balance Data
    • Source
  • Credit Card Default Data
  • Fund Manager Data
  • Baseball Data
    • Notes
  • Khan Gene Data
    • Format
    • Notes
  • NCI 60 Data
    • Source
  • New York Stock Exchange Data
    • Source
  • Orange Juice Data
    • Source
  • Portfolio Data
  • Time-to-Publication Data
    • Source
  • S&P Stock Market Data
    • Source
  • Violent Crime Rates by US State
    • Notes
  • Mid-Atlantic Wage Data
    • Source
  • Weekly S&P Stock Market Data
    • Source
• Transforms for flexible features
  • Polynomial features
    • Compare to R
    • Underlying model
    • Raw polynomials
    • Intercept
  • Spline features
    • Compare to R
    • Underlying model
    • Natural splines
    • Intercept
  • Derived features: using PCA on a subset of columns
    • Helper function
    • Direct comparison
• Tools for regression models
  • Building design matrices with ModelSpec
    • Interactions
    • Ordinal variables
    • Structure of a ModelSpec
      • Feature objects
    • Predicting at new points
    • Using np.ndarray
      • Caveats using np.ndarray
  • Model selection using ModelSpec
    • Forward Selection
  • ANOVA using ModelSpec
    • Grouping variables
• Helper functions
  • Clustering
    • Make a toy dataset
    • Cluster it
    • Plot the dendrogram
  • Generalized Additive Models
    • Make a toy dataset
    • Create a GAM
    • Plot the partial dependence plot for first feature
    • Fixing degrees of freedom
      • Create a new GAM with the correctly fixed terms
  • Survival Analysis
    • Define a cumulative hazard
    • Plot survival function
  • Support Vector Machines
    • Make a toy dataset
    • Fit an SVM classifier
    • Slicing through different features
• Labs
  • Package versions
    • Introduction to Python
      • Getting Started
      • Basic Commands
      • Introduction to Numerical Python
      • Graphics
      • Sequences and Slice Notation
      • Indexing Data
        • Indexing Rows, Columns, and Submatrices
        • Boolean Indexing
      • Loading Data
        • Reading in a Data Set
        • Basics of Selecting Rows and Columns
        • More on Selecting Rows and Columns
      • For Loops
        • String Formatting
      • Additional Graphical and Numerical Summaries
    • Linear Regression
      • Importing packages
        • New imports
        • Inspecting Objects and Namespaces
      • Simple Linear Regression
        • Using Transformations: Fit and Transform
        • Defining Functions
      • Multiple Linear Regression
      • Multivariate Goodness of Fit
        • List Comprehension
      • Interaction Terms
      • Non-linear Transformations of the Predictors
      • Qualitative Predictors
    • Logistic Regression, LDA, QDA, and KNN
      • The Stock Market Data
      • Logistic Regression
      • Linear Discriminant Analysis
      • Quadratic Discriminant Analysis
      • Naive Bayes
      • K-Nearest Neighbors
        • Tuning Parameters
        • Comparison to Logistic Regression
      • Linear and Poisson Regression on the Bikeshare Data
        • Linear Regression
        • Poisson Regression
    • Cross-Validation and the Bootstrap
      • The Validation Set Approach
      • Cross-Validation
      • The Bootstrap
        • Estimating the Accuracy of a Statistic of Interest
        • Estimating the Accuracy of a Linear Regression Model
    • Linear Models and Regularization Methods
      • Subset Selection Methods
        • Forward Selection
        • Choosing Among Models Using the Validation Set Approach and Cross-Validation
        • Best Subset Selection
      • Ridge Regression and the Lasso
        • Ridge Regression
        • Estimating Test Error of Ridge Regression
        • Fast Cross-Validation for Solution Paths
        • Evaluating Test Error of Cross-Validated Ridge
        • The Lasso
      • PCR and PLS Regression
        • Principal Components Regression
        • Partial Least Squares
    • Non-Linear Modeling
      • Polynomial Regression and Step Functions
      • Splines
      • Smoothing Splines and GAMs
        • Additive Models with Several Terms
        • ANOVA Tests for Additive Models
      • Local Regression
    • Tree-Based Methods
      • Fitting Classification Trees
      • Fitting Regression Trees
      • Bagging and Random Forests
      • Boosting
      • Bayesian Additive Regression Trees
    • Support Vector Machines
      • Support Vector Classifier
      • Support Vector Machine
      • ROC Curves
      • SVM with Multiple Classes
      • Application to Gene Expression Data
    • Deep Learning
      • Torch-Specific Imports
      • Single Layer Network on Hitters Data
        • Linear Models
        • Specifying a Network: Classes and Inheritance
        • Cleanup
      • Multilayer Network on the MNIST Digit Data
      • Convolutional Neural Networks
        • Hardware Acceleration
      • Using Pretrained CNN Models
      • IMDB Document Classification
        • Comparison to Lasso
      • Recurrent Neural Networks
        • Sequential Models for Document Classification
        • Time Series Prediction
    • Survival Analysis
      • Brain Cancer Data
      • Publication Data
      • Call Center Data
    • Unsupervised Learning
      • Principal Components Analysis
      • Matrix Completion
      • Clustering
        • $K$-Means Clustering
        • Hierarchical Clustering
      • NCI60 Data Example
        • PCA on the NCI60 Data
        • Clustering the Observations of the NCI60 Data
    • Multiple Testing
      • Review of Hypothesis Tests
      • Family-Wise Error Rate
      • False Discovery Rate
      • A Re-Sampling Approach
• Creating IMDB dataset from keras version
  • Store as sparse tensors
    • Save as sparse scipy matrices
  • Save and pickle the word index
  • Padded representations