fluidsim.util.console.bench_analysis

Load and plot benchmarks (fluidsim.util.console.bench_analysis)

Functions

exit_if_empty(df, input_params)	Check if the dataframe is empty.
filter_by_shape(df, n0, n1[, n2])	Filters all results with the same n0 and same n1.
filter_by_shapeloc(df, n0_loc, n1_loc[, n2_loc])	Filters all results with the same n0_loc * n1_loc.
init_parser(parser)	Initialize argument parser for fluidsim bench-analysis.
load_bench(path_dir, solver[, hostname])	Load benchmarks results saved as JSON files.
plot_scaling(path_dir, solver, hostname, n0, n1)	Plot speedup vs number of processes from benchmark results.
run(args)	Run fluidsim bench-analysis command.