Proj2

Real Estate Prediction

Buenos Aires & Mexico City Price Analysis

Overview

Proj2 builds regression models to predict apartment prices in Buenos Aires and Mexico City. The project progressively adds features and complexity across 4 lessons.

Data Source

CSV files with apartment listings including size, location, and neighborhood.

Techniques

Linear Regression, Ridge, Feature Engineering, OneHotEncoder

Evaluation

Mean Absolute Error (MAE), baseline comparison

Lesson Workflows

Lesson Topic Pipeline Query
2.1 Price and Size Size -> LinearRegression "run lesson 2.1"
2.2 Price and Location Lat/Lon -> Imputer + LinearRegression "run lesson 2.2"
2.3 Price and Neighborhood Neighborhood -> OneHotEncoder + Ridge "run lesson 2.3"
2.4 Price and Everything All Features -> Full Pipeline "run lesson 2.4"

Data Files

Proj2/data/ ├── buenos-aires--1.csv # Lesson 2.1 data ├── buenos-aires--2.csv # Lesson 2.2 data ├── buenos-aires--3.csv # Lesson 2.3 data ├── buenos-aires--4.csv # Lesson 2.4 data ├── buenos-aires--5.csv # Lesson 2.5 assignment ├── buenos-aires--test.csv # Test data ├── mexico-city--*.csv # Parallel datasets └── cache/ # Parquet cache

Available Tools

Tool Description
load_proj2_data Load Buenos Aires or Mexico City dataset
analyze_real_estate Exploratory analysis with statistics
predict_price Run price prediction pipeline
wrangle_data Apply lesson-specific wrangling
create_mapbox Visualize locations on interactive map

Quick Start

# Load Buenos Aires data "load buenos aires data version 1" # Run complete lesson "run lesson 2.1" # Analyze with specific city "analyze real estate in mexico city" # Compare cities "compare buenos aires and mexico city prices"