🏠 California Housing Dataset

Generated California Housing dataset files for machine learning projects.

📥 Download Individual Files

california_housing.csv (~1.8MB, 20,640 rows)
california_housing_train.csv (~1.8MB, 16,512 rows)
california_housing_test.csv (~449KB, 4,128 rows)

🤖 Pre-trained Model

trained_model.joblib (LinearRegression Model)

📦 Complete Package (Data + Source Code)

For the complete course materials including Python source code:

🚀 Download Latest Release

Includes data files, training scripts, and dependencies

🔗 Quick Access URLs

# Data files
wget https://codesignal-learn.github.io/course_building-aws-sagemaker/california_housing_train.csv
wget https://codesignal-learn.github.io/course_building-aws-sagemaker/california_housing_test.csv

# Pre-trained model
wget https://github.com/CodeSignal-Learn/course_building-aws-sagemaker/releases/latest/download/trained_model.joblib

📊 Dataset Information

Source: Scikit-learn California Housing Dataset

Target: Median house value prediction

Split: 80% train / 20% test (random_state=42)

Features

Original (9 features): MedInc, HouseAge, AveRooms, AveBedrms, Population, AveOccup, Latitude, Longitude, MedHouseVal
Train/Test (10 features): Original features + RoomsPerHousehold (engineered feature)
Preprocessing: Values capped at 95th percentile (excluding coordinates)