Preparing a dataset to estimate discrete choice models

The dataset prepared for these recipes consists of 10,000 choice situations. The first column contains the ID of the chosen alternative and the second column contains a list of all the alternatives the person was choosing from; not all of the alternatives were available in all choice situations.

Getting ready

To execute this query, you will need pandas, NumPy, and the regular expressions module. No other prerequisites are required.

How to do it…

Python Biogeme requires a tab-separated dataset, where each row contains the attributes of all the alternatives and the flag for the chosen one (an integer) as well as an indicator whether the alternative was available to the decision-maker at the time when the decision was made. Only numerical values are allowed. The following code can be found in the dcm_dataPrep.py file:

import pandas as pd
import numpy as np
import re

# read datasets
observations_filename = '../../Data/Chapter10/observations...