adelie.diagnostic.plot_benchmark#

adelie.diagnostic.plot_benchmark(total_time: ndarray, benchmark_screen: ndarray, benchmark_fit_screen: ndarray, benchmark_fit_active: ndarray, benchmark_kkt: ndarray, benchmark_invariance: ndarray, n_valid_solutions: ndarray, lmdas: ndarray, *, relative: bool = False)[source]#

Plots benchmark times.

Parameters:
total_timefloat

Total time taken for the core routine.

benchmark_screen(B,) ndarray

Benchmark timings for screening.

benchmark_fit_screen(B,) ndarray

Benchmark timings for fitting on screen set.

benchmark_fit_active(B,) ndarray

Benchmark timings for fitting on active set.

benchmark_kkt(B,) ndarray

Benchmark timings for KKT checks.

benchmark_invariance(B,) ndarray

Benchmark timings for invariance step.

n_valid_solutions(B,) ndarray

Flags that indicate whether each iteration resulted in a valid solution.

lmdas(L,) ndarray

Regularization parameters \(\lambda\).

relativebool, optional

If True, the time breakdown plot is relative to the total time, therefore plotting the proportion of time spent in each category. Otherwise, the absolute times are shown. Default is False.

Returns:
fig, ax