Skip to main content

Local Simulation

Use local simulation when you want to try FedGWAS on one machine. This mode starts multiple virtual clients through Flower local-simulation, so it is the fastest path for installation checks, smoke tests, and example runs.

Install​

Create and activate a Python environment, then install FedGWAS:

python -m pip install FedGWAS

Confirm that the simulation CLI is available:

fedgwas-sim --help

For full GWAS runs, install PLINK and make sure the executable is available on PATH:

plink --help
plink2 --help

Create a study​

Start from an empty directory owned by the user:

mkdir my_study
cd my_study

fedgwas-sim init
fedgwas-sim setup-experiment syn-tiny --seed 42

This writes a local Flower project, center configs, generated synthetic PLINK data, and output directories under my_study/.

Validate and run​

Run readiness checks before starting Flower:

fedgwas-sim check

Start the local simulation:

fedgwas-sim run --rounds 100

Evaluate outputs​

Generate a centralized comparison baseline, evaluate the federated outputs, and collect run metadata:

fedgwas-sim baseline generate --output data/centralized_baseline
fedgwas-sim evaluate results --baseline data/centralized_baseline --king
fedgwas-sim results collect --label tiny_run

Next steps​

  • Use fedgwas-sim CLI for the complete command reference.
  • Use Simulation Mode for repository-level Flower settings and runtime behavior.
  • Use Examples for correctness, performance, and 1000 Genomes scenarios.