Benchopt Benchmark for the Lasso Path

Benchopt is a package to simplify and make more transparent and reproducible the comparisons of optimization algorithms. This benchmark is dedicated to benchmarking algorithms that solve the full lasso path, that is, solving

$$\min_w \left(\frac{1}{2} \|y - Xw\|^2_2 + \lambda \|w\|_1\right)$$

for a sequence of $\lambda$ values.

Install

This benchmark can be run using the following commands:

conda create -n benchopt python
conda activate benchopt
pip install benchopt
git clone https://github.com/benchopt/benchmark_lasso_path
benchopt install -e benchmark_lasso_path
benchopt run -e benchmark_lasso_path

Apart from the problem, options can be passed to benchopt run, to restrict the benchmarks to some solvers or datasets, e.g.:

benchopt run benchmark_lasso_path -s celer -d simulated --max-runs 10 --n-repetitions 10

Use benchopt run -h for more details about these options, or visit https://benchopt.github.io/api.html.