Optimal Product Pricing

This package contains code for solving product pricing problems (PPPs) as described in our note on optimal product pricing (equation (2)). Three methods for solving a PPP are provided: quadratic minorization maximization (QMM), convex concave procedure (CCP), and nonlinear programming (NLP). For mathematical background, please refer to the manuscript.

Dependencies

This code depends on the Python packages CVXPY, SciPy, and Matplotlib (for figure generation). If you wish to use NLP, please install CVXPY from sources from this fork, until CVXPY with support for NLP is released.

Example

Here's a simple example for optimizing the prices of 1280 products with QMM:

from data import generate_data
from optimization import solve_ppp


# Generate sample data for 1280 products
profit_data, constraint_data = generate_data(n=1280)

# Solve using QMM
result = solve_ppp(profit_data, constraint_data, method='QMM')

# Inspect final profit and price changes
print(result.profit)
print(result.price_changes)

Results in manuscript

To reproduce the figures in the manuscript, run:

python figure2.py
python figure3.py