Skip to content

Commit

Permalink
MP instead of MT in experiments code
Browse files Browse the repository at this point in the history
  • Loading branch information
@kylebd99
kylebd99 committed Dec 20, 2023
1 parent 62bcf89 commit 440a764
Show file tree
Hide file tree
Showing 5 changed files with 41 additions and 32 deletions.
15 changes: 15 additions & 0 deletions Experiments/Experiments.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,9 @@ using DelimitedFiles: writedlm
using BenchmarkTools
using Random
using Printf
using SharedArrays
using WeakRefStrings
using Distributed

include("../Source/CardinalityWithColors.jl")
include("utils.jl")
Expand All @@ -18,3 +21,15 @@ include("build_color_summaries.jl")
include("get_true_cardinalities.jl")
include("run_estimators.jl")
include("graph_results.jl")
@everywhere include("../Source/CardinalityWithColors.jl")
@everywhere include("utils.jl")
@everywhere include("load_datasets.jl")
@everywhere include("build_color_summaries.jl")
@everywhere include("run_estimators.jl")
@everywhere using SharedArrays
@everywhere using WeakRefStrings
@everywhere using DelimitedFiles: writedlm
@everywhere using Parquet2: Dataset
@everywhere using Random
@everywhere using CSV, DataFrames
@everywhere using Serialization: serialize, deserialize
40 changes: 17 additions & 23 deletions Experiments/Scripts/comparison_exps.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,8 +3,8 @@ using Profile
include("../Experiments.jl")

#datasets = [human, aids, lubm80, yeast, hprd, dblp, youtube, eu2005, patents, wordnet]
datasets = [human, aids, lubm80, yeast, dblp, youtube, eu2005, patents]
#datasets = [aids]
#datasets = [human, aids, lubm80, yeast, dblp, youtube, eu2005, patents]
datasets = [human]

experiment_params = Vector{ExperimentParams}()
for dataset in datasets
Expand All @@ -16,17 +16,22 @@ for dataset in datasets
dataset=dataset,
partitioning_scheme=[(QuasiStable, 32), (NeighborNodeLabels, 32),(QuasiStable, 32), (NeighborNodeLabels, 32)],
description = "AvgQ64N64"))

push!(experiment_params, ExperimentParams(deg_stats_type=MinDegStats,
dataset=dataset,
partitioning_scheme=[(QuasiStable, 64)],
max_cycle_size = -1,
description = "MinQ64"))
push!(experiment_params, ExperimentParams(deg_stats_type=MaxDegStats,
dataset=dataset,
partitioning_scheme=[(QuasiStable, 64)],
max_cycle_size = -1,
description = "MaxQ64"))

push!(experiment_params, ExperimentParams(deg_stats_type=MaxDegStats,
dataset=dataset,
partitioning_scheme=[(Hash, 64)],
max_cycle_size = -1,
description = "BSK"))

push!(experiment_params, ExperimentParams(deg_stats_type=AvgDegStats,
Expand All @@ -36,9 +41,9 @@ for dataset in datasets
description = "IndEst"))
end

build_experiments(experiment_params)
#build_experiments(experiment_params)

#run_estimation_experiments(experiment_params)
run_estimation_experiments(experiment_params)

graph_grouped_boxplot_with_comparison_methods(experiment_params;
ylims=[10^-5, 10^4],
Expand All @@ -47,30 +52,19 @@ graph_grouped_boxplot_with_comparison_methods(experiment_params;
grouping=description,
dimensions = (1450, 550),
legend_pos=:top,
y_label="Runtime (10^ s)",
filename="runtime")
y_label="Inference Latency (10^ s)",
filename="overall_runtime")

graph_grouped_boxplot_with_comparison_methods(experiment_params;
ylims=[10^-21, 10^21],
y_ticks=[10^-20, 10^-15, 10^-10, 10^-5, 10^-2, 10^0, 10^2, 10^5, 10^10, 10^15, 10^20],
y_type = estimate_error,
grouping=description,
dimensions = (1450, 550),
legend_pos=:bottomleft,
legend_pos=:top,
y_label="Relative Error (10^)",
filename="error")

filename="overall_error")

graph_grouped_boxplot_with_comparison_methods(experiment_params;
ylims=[10^-21, 10^21],
x_type = query_size,
y_ticks=[10^-20, 10^-15, 10^-10, 10^-5, 10^-2, 10^0, 10^2, 10^5, 10^10, 10^15, 10^20],
y_type = estimate_error,
grouping=description,
dimensions = (1450, 550),
legend_pos=:bottomleft,
y_label="Relative Error (10^)",
filename="error-query-size")

graph_grouped_bar_plot(experiment_params;
grouping=description,
Expand All @@ -79,13 +73,13 @@ graph_grouped_bar_plot(experiment_params;
y_ticks = [5, 10, 15, 20, 25, 30],
dimensions = (1000, 550),
y_label="Memory (MBs)",
filename="memory")
filename="overall_memory")

graph_grouped_bar_plot(experiment_params;
grouping=description,
y_type=build_time,
ylims=[0, 100],
y_ticks = [20, 40, 60, 80, 100],
ylims=[0, 1600],
y_ticks = [200, 400, 600, 800, 1000, 1200, 1400, 1600],
dimensions = (1000, 550),
y_label="Build Time (s)",
filename="build_time")
filename="overall_build_time")
2 changes: 1 addition & 1 deletion Experiments/build_color_summaries.jl
View file
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
function build_experiments(experiment_params_list::Vector{ExperimentParams})
for experiment_params in experiment_params_list
@sync @distributed for experiment_params in experiment_params_list
build_times = [("Dataset", "Partitioner", "NumColors", "BuildPhase", "BuildTime", "MemoryFootprint")]
dataset = experiment_params.dataset
summary_params = experiment_params.summary_params
Expand Down
14 changes: 6 additions & 8 deletions Experiments/run_estimators.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,10 +5,8 @@ function run_estimation_experiments(experiment_params_list::Vector{ExperimentPar
summary_file_location = "Experiments/SerializedSummaries/" * params_to_summary_filename(experiment_params)
!isfile(summary_file_location) && error("The summary has not been built yet! \n Attempted File Location: $(summary_file_location)")
summary::ColorSummary = deserialize(summary_file_location)
experiment_results = []
push!(experiment_results, ("Estimate", "TrueCard", "EstimationTime", "QueryType", "QueryPath", "QuerySize"))
lk = ReentrantLock()
Threads.@threads for i in shuffle(collect(1:length(all_queries[dataset])))
experiment_results = SharedArray{Tuple{Float64, Float64, Float64, String255, String255, Float64}}(length(all_queries[dataset]))
@sync @distributed for i in 1:length(experiment_results)
query::QueryGraph = all_queries[dataset][i].query
query_path = all_queries[dataset][i].query_path
exact_size = all_queries[dataset][i].exact_size
Expand All @@ -20,11 +18,11 @@ function run_estimation_experiments(experiment_params_list::Vector{ExperimentPar
estimate_time = median([x.time for x in estimate_results]) # Convert back to seconds from nano seconds
estimate = max(1, estimate_results[1].value)
query_type = all_queries[dataset][i].query_type
lock(lk)
push!(experiment_results, (estimate, exact_size, estimate_time, query_type, query_path, nv(query.graph)))
unlock(lk)
experiment_results[i] = (estimate, exact_size, estimate_time, query_type, query_path, nv(query.graph))
end
final_results = [(x[1], x[2], x[3], String(x[4]), String(x[5]), x[6]) for x in experiment_results]
final_results = [("Estimate", "TrueCard", "EstimationTime", "QueryType", "QueryPath", "QuerySize"); final_results]
results_file_location = "Experiments/Results/Estimation_" * params_to_results_filename(experiment_params)
writedlm(results_file_location, experiment_results, ",")
writedlm(results_file_location, final_results, ",")
end
end
2 changes: 2 additions & 0 deletions Project.toml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -18,8 +18,10 @@ Probably = "2172800d-0309-5a57-a84f-d50c94757422"
QuasiStableColors = "9c3856af-3e7c-4d34-a6af-a406867b22e4"
Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
Revise = "295af30f-e4ad-537b-8983-00126c2a3abe"
SharedArrays = "1a1011a3-84de-559e-8e89-a11a2f7dc383"
StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
StatsPlots = "f3b207a7-027a-5e70-b257-86293d7955fd"
WeakRefStrings = "ea10d353-3f73-51f8-a26c-33c1cb351aa5"

[compat]
julia = "^1.6"
Expand Down

0 comments on commit 440a764

Please sign in to comment.