fix charge calc, add all cols and write to csv

results/data/actual-prices-binding/rev_and_negrev_standard_vs_apb.csv

@@ -0,0 +1,8 @@
+formulation,min_rev_improvements_lkgt5,med_rev_improvements,max_rev_improvement,max_neg_rev_discharge_standard,max_neg_rev_charge_500_standard,max_neg_rev_discharge_apb,max_neg_rev_charge_500_apb
+arbitrage_discounted_no_degradation/hyp,0.1445991015030699,0.291237702693409,0.4076671585964029,0.001114444487147587,0.28943078143789214,0.0009053806106194061,0.05221042374504637
+arbitrage_throughputpenalty_no_degradation/600000.0,0.13059840125960517,0.44107559482537595,0.7836292288023413,0.01135314258357969,0.36006635513280216,0.006962763887358539,0.05452351938234726
+arbitrage_throughputpenalty_no_degradation/800000.0,0.13706614001832815,0.45373874117785984,0.860130227419397,0.0015623148110588928,0.36129979787074185,0.001687379856623454,0.057291445453499935
+arbitrage_throughputpenalty_no_degradation/200000.0,0.11432871738825902,0.3226213156035771,0.683313688773886,0.01248551005727752,0.26975355037234633,0.007121915971583968,0.04314123218179427
+arbitrage_no_degradation,0.10096500362145246,0.27379020257143993,0.4978681272068267,0.03854378949162245,0.20499280215435836,0.023421797153251253,0.035830828982764774
+arbitrage_throughputpenalty_no_degradation/400000.0,0.1173053968329046,0.3718434366142929,0.7419037858184889,0.011516613236363781,0.34779621246503317,0.006874560238074517,0.05341673038601346
+arbitrage_discounted_no_degradation/exp,0.12139027359881101,0.2678396036469154,0.38753467396529745,0.0028037920512216205,0.2892076390500374,0.0034191761483650233,0.052496665018914285

results/standard_vs_apb_results.jl

@@ -2,6 +2,7 @@ using DataFrames
 using JLD2
 using NEMStorageUnderUncertainty
 using Statistics
+using CSV
 
 const SCENARIO_COLS = [
     :formulation, :energy_capacity, :power_capacity, :data_type, :lookahead
@@ -16,6 +17,8 @@ function merge_summary_data(standard_data_path::String, apb_data_path::String)
     all_merged = DataFrame[]
     for key in keys(standard_data)
         (df_standard, df_apb) = (standard_data[key], apb_data[key])
+        df_standard.formulation = fill(key, size(df_standard, 1))
+        df_apb.formulation = fill(key, size(df_apb, 1))
         df_standard_forecast = df_standard[df_standard.data_type.=="forecast", :]
         df_apb_forecast = df_apb[df_apb.data_type.=="forecast", :]
         merged = innerjoin(
@@ -32,22 +35,32 @@ end
 """
 Calculates three metrics
 
-- `rev_improvement`: Improvement in actual-prices-binding scenario
-    from standard scenario as % of standard revenue
-- `neg_rev_discharge_percent`: Negative revenue from discharge in standard scenario
+- `neg_rev_discharge_percent_standard`: Negative revenue from discharge in standard scenario
     as % of standard revenue
-- `neg_rev_charge_500_percent`: Negative revenue from charging in standard scenario
+- `neg_rev_charge_500_percent_standard`: Negative revenue from charging in standard scenario
     when price ≥ 500 as % of standard revenue
+- `neg_rev_discharge_percent_apb`: Negative revenue from discharge in APB scenario
+    as % of APB revenue
+- `neg_rev_charge_500_percent_apb`: Negative revenue from charging in APB scenario
+    when price ≥ 500 as % of APB revenue
+- `rev_improvement`: Improvement in actual-prices-binding scenario
+    from standard scenario as % of standard revenue
 """
 function calculate_metrics(all_merged_data::Array{DataFrame})
     all_metrics = DataFrame[]
     metric_cols = [
-        :rev_improvement_percent, :neg_rev_discharge_percent, :neg_rev_charge_500_percent
+        :rev_improvement,
+        :neg_rev_discharge_standard,
+        :neg_rev_charge_500_standard,
+        :neg_rev_discharge_apb,
+        :neg_rev_charge_500_apb,
     ]
     for df in all_merged_data
-        df[:, metric_cols[1]] = @. (df[:, :revenue_apb] - df[:, :revenue_standard]) / (df[:, :revenue_standard]) * 100
-        df[:, metric_cols[2]] = @. -1 * df[:, :neg_rev_discharge_standard] / (df[:, :revenue_standard]) * 100
-        df[:, metric_cols[3]] = @. -1 * df[:, :neg_rev_discharge_standard] / (df[:, :revenue_standard]) * 100
+        df[:, metric_cols[1]] = @. (df[:, :revenue_apb] - df[:, :revenue_standard]) / (df[:, :revenue_standard])
+        df[:, metric_cols[2]] = @. -1 * df[:, :neg_rev_discharge_standard] / (df[:, :revenue_standard])
+        df[:, metric_cols[3]] = @. -1 * df[:, :neg_rev_charge_500_standard] / (df[:, :revenue_standard])
+        df[:, metric_cols[4]] = @. -1 * df[:, :neg_rev_discharge_apb] / (df[:, :revenue_apb])
+        df[:, metric_cols[5]] = @. -1 * df[:, :neg_rev_charge_500_apb] / (df[:, :revenue_apb])
         push!(all_metrics, df[:, cat(SCENARIO_COLS, metric_cols, dims=1)])
     end
     return all_metrics
@@ -62,27 +75,34 @@ function summarise_metrics(all_metrics::Array{DataFrame})
     min_rev_improvements = Float64[]
     med_rev_improvements = Float64[]
     max_rev_improvements = Float64[]
-    max_neg_rev_discharge_percents = Float64[]
-    max_neg_rev_charge_500_percents = Float64[]
+    max_neg_rev_discharge_standard = Float64[]
+    max_neg_rev_charge_500_standard = Float64[]
+    max_neg_rev_discharge_apb = Float64[]
+    max_neg_rev_charge_500_apb = Float64[]
     for df in all_metrics
         filtered_df = _filter_results!(df)
         push!(formulations, only(unique(filtered_df.formulation)))
-        push!(min_rev_improvements, minimum(filtered_df[filtered_df.lookahead.!="5", :rev_improvement_percent]))
-        push!(med_rev_improvements, median(filtered_df.rev_improvement_percent))
-        push!(max_rev_improvements, maximum(filtered_df.rev_improvement_percent))
-        push!(max_neg_rev_discharge_percents, maximum(filtered_df.neg_rev_discharge_percent))
-        push!(max_neg_rev_charge_500_percents, maximum(filtered_df.neg_rev_charge_500_percent))
+        push!(min_rev_improvements, minimum(filtered_df[filtered_df.lookahead.!="5", :rev_improvement]))
+        push!(med_rev_improvements, median(filtered_df.rev_improvement))
+        push!(max_rev_improvements, maximum(filtered_df.rev_improvement))
+        push!(max_neg_rev_discharge_standard, maximum(filtered_df.neg_rev_discharge_standard))
+        push!(max_neg_rev_charge_500_standard, maximum(filtered_df.neg_rev_charge_500_standard))
+        push!(max_neg_rev_discharge_apb, maximum(filtered_df.neg_rev_discharge_apb))
+        push!(max_neg_rev_charge_500_apb, maximum(filtered_df.neg_rev_charge_500_apb))
     end
     return DataFrame(
         :formulation => formulations,
         :min_rev_improvements_lkgt5 => min_rev_improvements,
         :med_rev_improvements => med_rev_improvements,
         :max_rev_improvement => max_rev_improvements,
-        :max_neg_rev_discharge_percent => max_neg_rev_discharge_percents,
-        :max_neg_rev_charge_500_percent => max_neg_rev_charge_500_percents
+        :max_neg_rev_discharge_standard => max_neg_rev_discharge_standard,
+        :max_neg_rev_charge_500_standard => max_neg_rev_charge_500_standard,
+        :max_neg_rev_discharge_apb => max_neg_rev_discharge_apb,
+        :max_neg_rev_charge_500_apb => max_neg_rev_charge_500_apb,
     )
 end
 
 merged = merge_summary_data("results/data/NSW_summary_results.jld2", "results/data/actual-prices-binding/NSW_summary_results_apb.jld2")
 metrics = calculate_metrics(merged)
 summary = summarise_metrics(metrics)
+CSV.write("results/data/actual-prices-binding/rev_and_negrev_standard_vs_apb.csv", summary)