add query size graphs

Experiments/Experiments.jl

@@ -8,6 +8,7 @@ using Parquet2: Dataset
 using DelimitedFiles: writedlm
 using BenchmarkTools
 using Random
+using Printf
 
 include("../Source/CardinalityWithColors.jl")
 include("utils.jl")

Experiments/Scripts/query_size_exps.jl

@@ -0,0 +1,69 @@
+
+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]
+
+experiment_params = Vector{ExperimentParams}()
+for dataset in datasets
+    push!(experiment_params, ExperimentParams(deg_stats_type=AvgDegStats,
+                                                dataset=dataset,
+                                                partitioning_scheme=[(QuasiStable, 64)],
+                                                description = "AvgQ64"))
+    push!(experiment_params, ExperimentParams(deg_stats_type=AvgDegStats,
+                                                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)],
+                                                description = "MinQ64"))
+
+    push!(experiment_params, ExperimentParams(deg_stats_type=MaxDegStats,
+                                                dataset=dataset,
+                                                partitioning_scheme=[(QuasiStable, 64)],
+                                                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,
+                                                dataset=dataset,
+                                                partitioning_scheme=[(QuasiStable, 1)],
+                                                max_cycle_size = -1,
+                                                description = "IndEst"))
+end
+
+#build_experiments(experiment_params)
+
+#run_estimation_experiments(experiment_params)
+
+graph_grouped_boxplot_with_comparison_methods(experiment_params;
+                                                ylims=[10^-5, 10^4],
+                                                y_ticks=[10^-5, 10^-2, 10^0, 10^2, 10^4],
+                                                x_type = query_size,
+                                                y_type = runtime,
+                                                grouping=description,
+                                                dimensions = (1450, 550),
+                                                legend_pos=:topleft,
+                                                y_label="Runtime 10^ (s)",
+                                                x_label = "Query Size",
+                                                filename="query_size_runtime")
+
+
+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^",
+                                                x_label = "Query Size",
+                                                filename="query_size_error")

Experiments/graph_results.jl

@@ -107,6 +107,15 @@ function get_query_id(dataset, query_path)
     end
 end
 
+function query_size_category(s)
+    categories = [3, 4, 6, 9, 12, 16, 24, 32]
+    for cat in categories
+        if s <= cat
+            return cat
+        end
+    end
+end
+
 function graph_grouped_boxplot_with_comparison_methods(experiment_params_list::Vector{ExperimentParams};
                                         x_type::GROUP=dataset,
                                         y_type::VALUE=estimate_error,
@@ -136,7 +145,7 @@ function graph_grouped_boxplot_with_comparison_methods(experiment_params_list::V
             current_x = if x_type == dataset
                 data
             elseif x_type == query_size
-                results_df[i, :QuerySize]
+                @sprintf "%02i" query_size_category(results_df[i, :QuerySize])
             end
             current_group = string(grouping == query_type ? results_df[i, :QueryType] : get_value_from_param(experiment_params, grouping))
             current_y = if y_type == estimate_error
@@ -146,7 +155,7 @@ function graph_grouped_boxplot_with_comparison_methods(experiment_params_list::V
             end
             true_card[(data, get_query_id(string(experiment_params.dataset), results_df[i, :QueryPath]))] = (results_df[i, :TrueCard], current_x)
             # push the errors and their groupings into the correct vector
-            push!(x_values, current_x)
+            push!(x_values, string(current_x))
             push!(y_values, current_y)
             push!(estimators, current_group)
         end
@@ -179,13 +188,16 @@ function graph_grouped_boxplot_with_comparison_methods(experiment_params_list::V
                 runtime / 1000.0
             end
             # push the errors and their groupings into the correct vector
-            push!(x_values, current_x)
+            push!(x_values, string(current_x))
             push!(y_values, current_y)
             push!(estimators, estimator)
         end
     end
+    sorted_vals = sort(zip(x_values, y_values, estimators), by=(x)->x[1])
+    x_values = [x[1] for x in sorted_vals]
+    y_values = [x[2] for x in sorted_vals]
+    estimators = [x[3] for x in sorted_vals]
     println("starting graphs")
-
     # This seems to be necessary for using Plots.jl outside of the ipynb framework.
     # See this: https://discourse.julialang.org/t/deactivate-plot-display-to-avoid-need-for-x-server/19359/15
     ENV["GKSwstype"]="100"
@@ -196,7 +208,7 @@ function graph_grouped_boxplot_with_comparison_methods(experiment_params_list::V
                             y_ticks = [log10(y) for y in y_ticks],
                             legend = legend_pos,
                             size = dimensions,
-                            bottom_margin = 20px,
+                            bottom_margin = 40px,
                             top_margin = 20px,
                             left_margin = 10mm,
                             legend_column = 2,