This folder contains the simulated data used in the Management Science paper "A Bayesian Semiparametric Approach for Endogeneity and Heterogeneity in Choice Models" by Yang Li and Asim Ansari. 

Here are a list of data files and their usages:

-- "choice_new.dat" contains the identifiers of the chosen brands, ordered first by consumer and then by time period. '0' indicates the first brand, and the largest integer indicates the outside option.

-- "P.dat" contains prices of all choice occasions, ordered first by time period and then by choice alternative.

-- "X.dat" contains brand dummies and the exogenous variables, ordered first by time period and then by choice alternative.

-- "Z.dat" contains instruments, ordered first by time period and then by choice alternative.

-- "Index.dat", "Index1.dat", "Index2.dat", "Index4.dat" are indices files that facilitate matrix assembling and computation in the estimation algorithm.


For more details, please refer to the paper or contact Yang Li at yangli@ckgsb.edu.cn.