pnne-search 0.1.2

A package for search model estimation

This package pnne_search implements the pre-trained neural network estimator for sequential search model, as described by "Pre-Training Estimators for Structural Models: Application to Consumer Search"

See below for a simple demonstration of how to use pnne_search:

import time import pnne_search

data = pnne_search.load_example_data()

pnne_search.pnne_estimate(data['Y'], data['Xp'], data['Xa'], data['Xc'], data['consumer_idx'], checks = True)

start_time = time.time() pnne_search.pnne_estimate(data['Y'], data['Xp'], data['Xa'], data['Xc'], data['consumer_idx'], checks = True, se=True) time.time() - start_time

start_time = time.time() pnne_search.pnne_estimate(data['Y'], data['Xp'], data['Xa'], data['Xc'], data['consumer_idx'], checks = True, se=True, use_parallel=False) time.time() - start_time