Skip to content

Commit

Permalink
Update benchmark by returning outputs
Browse files Browse the repository at this point in the history
  • Loading branch information
Tortar authored and AldoGl committed Oct 28, 2024
1 parent 8b9f65c commit 591706a
Showing 1 changed file with 14 additions and 11 deletions.
25 changes: 14 additions & 11 deletions examples/benchmark_w_matlab.jl
Original file line number Diff line number Diff line change
@@ -1,28 +1,30 @@

# # Comparing the performance of the Julia and MATLAB implementations

# We can compare the performance of the Julia and MATLAB implementations by running the same model for the same number of epochs and measuring the time taken.
# We can compare the performance of the Julia and MATLAB implementations
# by running the same model for the same number of epochs and measuring
# the time taken.

using BeforeIT, Plots, Statistics

parameters = BeforeIT.AUSTRIA2010Q1.parameters
initial_conditions = BeforeIT.AUSTRIA2010Q1.initial_conditions
T = 12*2

# We will run the model without any output to avoid the overhead of printing the results.
function run_no_output(;multi_threading = false)
function run(; multi_threading = false)
model = BeforeIT.init_model(parameters, initial_conditions, T)
data = BeforeIT.init_data(model);

for _ in 1:T
BeforeIT.one_epoch!(model; multi_threading = multi_threading)
BeforeIT.update_data!(data, model)
end
return model, data
end

# we run the code to compile it fist
@time run_no_output()
@time run_no_output(;multi_threading = true)
# we run the code to compile it first
@time run();
@time run(;multi_threading = true);

# time taken by the MATLAB code, computed independently on an Apple M1 chip
matlab_times = [3.1919, 3.2454, 3.1501, 3.1074, 3.1551]
Expand All @@ -34,15 +36,14 @@ n_runs = 5

julia_times_1_thread = zeros(n_runs)
for i in 1:n_runs
julia_times_1_thread[i] = @elapsed run_no_output()
julia_times_1_thread[i] = @elapsed run();
end
julia_time_1_thread = mean(julia_times_1_thread)
julia_time_1_thread_std = std(julia_times_1_thread)


julia_times_multi_thread = zeros(n_runs)
for i in 1:5
julia_times_multi_thread[i] = @elapsed run_no_output(multi_threading = true)
julia_times_multi_thread[i] = @elapsed run(multi_threading = true);
end
julia_time_multi_thread = mean(julia_times_multi_thread)
julia_time_multi_thread_std = std(julia_times_multi_thread)
Expand All @@ -51,15 +52,17 @@ julia_time_multi_thread_std = std(julia_times_multi_thread)
n_threads = Threads.nthreads()

theme(:default, bg = :white)

# bar chart of the time taken vs the time taken by the MATLAB code, also plot the stds as error bars
# make a white background with no grid
bar(["MATLAB", "Julia, 1 thread", "Julia, $n_threads threads"], [matlab_time, julia_time_1_thread, julia_time_multi_thread],
yerr = [matlab_time_std, julia_time_1_thread_std, julia_time_multi_thread_std],
legend = false, dpi=300, size=(400, 300), grid = false, ylabel = "Time for one simulation (s)")

# the Julia implementation is faster than the MATLAB implementation, and the multi-threaded version is faster than the single-threaded version.
# the Julia implementation is faster than the MATLAB implementation, and the multi-threaded version is
# faster than the single-threaded version.

# increase

# save the image
savefig("benchmark_w_matlab.png")
savefig("benchmark_w_matlab.png")

0 comments on commit 591706a

Please sign in to comment.