Supplementary code to "Optimal Search and Discovery"

This Julia package replicates the results presented in Section 6.4 and produces simulations for a wider range of parameter values not presented in the paper.

Paper: https://doi.org/10.1287/mnsc.2021.4085

Dependencies

• Julia 1.5 - 1.6.5 (in 1.7 and later, the rng changed, hence using more recent Julia versions will not precisely reproduce the results)
• Packages specified in "Project.toml"

Installation instructions

1. Extract files from this package into a folder (e.g. via git clone https://github.com/rgreminger/SearchDiscoveryReplication.jl)
2. In Julia, navigate to the folder (cd("path/to/folder"))
3. Activate environment by running using Pkg; Pkg.activate(".")
4. Install packages by running Pkg.instantiate()

Overview main files

• replicate_main.jl: Produces results in Section 6.4 (results are saved in /gen)
• replicate_other_specs.jl: Produces results for other parameter values (note that this can take some time)
• replicate_test_estim.jl: Checks whether models recover parameters when data is generated from self
• src/replication.jl: Contains functions specific for replication

Documentation for the functions is provided through docstrings, which in Julia can be accessed through ?function (e.g. ?runSimulations).

Additional notes

Reservation values

The following document provides the additional derivations underlying two functions in src/reservation_values.jl: notes_reservation_values.pdf

(Github unfortunately does not render LaTex equations...)

Multi-threading

Several core functions can use multi-threading for considerable performance improvements. To use this, start Julia with multiple threads enabled (see Julia Docs).