QuantEcon.jl

For automated benchmarking and reporting, use PkgBenchmark.jl. This requires PkgBenchmark to be installed in your default environment.

To benchmark the QuantEcon package and export the results to a Markdown file:

using PkgBenchmark

results = benchmarkpkg("QuantEcon")
export_markdown("results.md", results)

You can access the official documentation and learning materials for the library through the following resources:

• Library Website: https://quantecon.org/quantecon-jl/
• Stable Documentation: https://quantecon.github.io/QuantEcon.jl/stable/
• Development Documentation: https://quantecon.github.io/QuantEcon.jl/dev/
• Lecture Site: The QuantEcon lecture site provides practical examples of the library in action.

You can evaluate performance changes between a target commit (or branch) and a baseline using the judge function in PkgBenchmark.jl.

Note: judge checks out and runs each commit separately; uncommitted changes in your working tree will not be included in the comparison.

To compare the current master branch against its previous commit (master~1):

jud = judge("QuantEcon", "master", "master~1")
export_markdown("judgement.md", jud)

To inspect the results in the REPL:

• View the judgment summary: show(PkgBenchmark.benchmarkgroup(jud))
• View baseline timing estimates: show(jud.baseline_results.benchmarkgroup)
• View target timing estimates: show(jud.target_results.benchmarkgroup)

After installation, you can load the package into your Julia session or script using the using keyword.

using QuantEcon

To install the QuantEcon package, open a Julia session and use the built-in package manager Pkg to add the package to your environment.

using Pkg
Pkg.add("QuantEcon")

Once the environment is set up, you can execute the entire benchmark suite using the benchmarks.jl entry point. This process typically takes several minutes.

julia --project=benchmark benchmark/benchmarks.jl

To install the QuantEcon package, use the Julia package manager (Pkg) within a Julia session. This will install the package via git to your default Julia library location.

using Pkg
Pkg.add("QuantEcon")

If you want to run only a specific subset of benchmarks (for example, a single kernel within a module), you can include the benchmark file in an interactive Julia session and access the SUITE object.

Example: Running the bellman_operator for the dense_n500_m100 case in the ddp module:

include("benchmark/benchmarks.jl")
run(SUITE["ddp"]["dense_n500_m100"]["bellman_operator"])

include("benchmark/benchmarks.jl");

run(SUITE["ddp"]["dense_n500_m100"]["bellman_operator"])

The benchmark suite is organized into a BenchmarkGroup named SUITE. The following modules are currently covered:

• SUITE["ddp"]: Benchmarks DiscreteDP (found in src/markov/ddp.jl).
• SUITE["lcp_lemke"]: Benchmarks lcp_lemke (found in src/lcp_lemke.jl).
• SUITE["mc_tools"]: Benchmarks MarkovChain (found in src/markov/mc_tools.jl).