Iteration 130: Add 8 benchmark pairs (364 total, +8 vs best 356)

Added benchmark pairs: skew_kurt, sem_var, mode_series, idxmin_idxmax,
dataframe_skew_kurt, dataframe_sem_var, nancumops, clip_advanced.

Run: https://github.com/githubnext/tsessebe/actions/runs/24528172385

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: github-actions[bot]

benchmarks/pandas/bench_clip_advanced.py

@@ -0,0 +1,33 @@
+"""
+Benchmark: Series.clip(lower_arr, upper_arr) / DataFrame.clip() — per-element clipping with array bounds.
+Outputs JSON: {"function": "clip_advanced", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import math
+import time
+import pandas as pd
+import numpy as np
+
+ROWS = 50_000
+WARMUP = 5
+ITERATIONS = 30
+
+data = np.array([math.sin(i * 0.01) * 200 for i in range(ROWS)])
+lower = np.full(ROWS, -50.0)
+upper = np.full(ROWS, 50.0)
+s = pd.Series(data)
+
+df_data = {f"col{c}": np.array([math.sin((i + c) * 0.01) * 200 for i in range(ROWS)]) for c in range(5)}
+df = pd.DataFrame(df_data)
+
+for _ in range(WARMUP):
+    s.clip(lower=lower, upper=upper)
+    df.clip(lower=-50, upper=50)
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    s.clip(lower=lower, upper=upper)
+    df.clip(lower=-50, upper=50)
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "clip_advanced", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/pandas/bench_dataframe_sem_var.py

@@ -0,0 +1,29 @@
+"""
+Benchmark: DataFrame.var() / DataFrame.sem() — variance and SEM on a 10k×10 DataFrame.
+Outputs JSON: {"function": "dataframe_sem_var", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import math
+import time
+import pandas as pd
+import numpy as np
+
+ROWS = 10_000
+COLS = 10
+WARMUP = 5
+ITERATIONS = 20
+
+data = {f"col{c}": np.array([math.sin((i + c) * 0.01) * 100 for i in range(ROWS)]) for c in range(COLS)}
+df = pd.DataFrame(data)
+
+for _ in range(WARMUP):
+    df.var()
+    df.sem()
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    df.var()
+    df.sem()
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "dataframe_sem_var", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/pandas/bench_dataframe_skew_kurt.py

@@ -0,0 +1,29 @@
+"""
+Benchmark: DataFrame.skew() / DataFrame.kurt() — skewness and kurtosis on a 10k×10 DataFrame.
+Outputs JSON: {"function": "dataframe_skew_kurt", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import math
+import time
+import pandas as pd
+import numpy as np
+
+ROWS = 10_000
+COLS = 10
+WARMUP = 5
+ITERATIONS = 20
+
+data = {f"col{c}": np.array([math.sin((i + c) * 0.01) * 100 for i in range(ROWS)]) for c in range(COLS)}
+df = pd.DataFrame(data)
+
+for _ in range(WARMUP):
+    df.skew()
+    df.kurt()
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    df.skew()
+    df.kurt()
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "dataframe_skew_kurt", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/pandas/bench_idxmin_idxmax.py

@@ -0,0 +1,28 @@
+"""
+Benchmark: Series.idxmin() / Series.idxmax() — index of min/max on a 100k-element Series.
+Outputs JSON: {"function": "idxmin_idxmax", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import math
+import time
+import pandas as pd
+import numpy as np
+
+SIZE = 100_000
+WARMUP = 5
+ITERATIONS = 50
+
+data = np.array([math.sin(i * 0.01) * 1000 for i in range(SIZE)])
+s = pd.Series(data)
+
+for _ in range(WARMUP):
+    s.idxmin()
+    s.idxmax()
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    s.idxmin()
+    s.idxmax()
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "idxmin_idxmax", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/pandas/bench_mode_series.py

@@ -0,0 +1,24 @@
+"""
+Benchmark: Series.mode() — mode of a 10k-element integer Series.
+Outputs JSON: {"function": "mode_series", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import time
+import pandas as pd
+
+SIZE = 10_000
+WARMUP = 5
+ITERATIONS = 50
+
+data = [i % 200 for i in range(SIZE)]
+s = pd.Series(data)
+
+for _ in range(WARMUP):
+    s.mode()
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    s.mode()
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "mode_series", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/pandas/bench_nancumops.py

@@ -0,0 +1,35 @@
+"""
+Benchmark: np.nansum / np.nanmean / np.nanvar / np.nanstd — nan-ignoring aggregates on 100k array.
+Outputs JSON: {"function": "nancumops", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import math
+import time
+import numpy as np
+
+SIZE = 100_000
+WARMUP = 5
+ITERATIONS = 50
+
+# Array with ~10% NaN values
+data = np.array([float("nan") if i % 10 == 0 else math.sin(i * 0.01) * 100 for i in range(SIZE)])
+
+for _ in range(WARMUP):
+    np.nansum(data)
+    np.nanmean(data)
+    np.nanvar(data)
+    np.nanstd(data)
+    np.nanmin(data)
+    np.nanmax(data)
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    np.nansum(data)
+    np.nanmean(data)
+    np.nanvar(data)
+    np.nanstd(data)
+    np.nanmin(data)
+    np.nanmax(data)
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "nancumops", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/pandas/bench_sem_var.py

@@ -0,0 +1,28 @@
+"""
+Benchmark: Series.var() / Series.sem() — variance and SEM on a 100k-element Series.
+Outputs JSON: {"function": "sem_var", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import math
+import time
+import pandas as pd
+import numpy as np
+
+SIZE = 100_000
+WARMUP = 5
+ITERATIONS = 50
+
+data = np.array([math.sin(i * 0.01) * 100 for i in range(SIZE)])
+s = pd.Series(data)
+
+for _ in range(WARMUP):
+    s.var()
+    s.sem()
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    s.var()
+    s.sem()
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "sem_var", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/pandas/bench_skew_kurt.py

@@ -0,0 +1,28 @@
+"""
+Benchmark: Series.skew() / Series.kurt() — skewness and kurtosis on a 100k-element Series.
+Outputs JSON: {"function": "skew_kurt", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+"""
+import json
+import math
+import time
+import pandas as pd
+import numpy as np
+
+SIZE = 100_000
+WARMUP = 5
+ITERATIONS = 50
+
+data = np.array([math.sin(i * 0.01) * 100 for i in range(SIZE)])
+s = pd.Series(data)
+
+for _ in range(WARMUP):
+    s.skew()
+    s.kurt()
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    s.skew()
+    s.kurt()
+total = (time.perf_counter() - start) * 1000
+
+print(json.dumps({"function": "skew_kurt", "mean_ms": total / ITERATIONS, "iterations": ITERATIONS, "total_ms": total}))

benchmarks/tsb/bench_clip_advanced.ts

@@ -0,0 +1,36 @@
+/**
+ * Benchmark: clipAdvancedSeries / clipAdvancedDataFrame — per-element clipping with array bounds.
+ * Outputs JSON: {"function": "clip_advanced", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+ */
+import { Series, DataFrame, clipAdvancedSeries, clipAdvancedDataFrame } from "../../src/index.ts";
+
+const ROWS = 50_000;
+const WARMUP = 5;
+const ITERATIONS = 30;
+
+const data = Float64Array.from({ length: ROWS }, (_, i) => Math.sin(i * 0.01) * 200);
+const lower = Float64Array.from({ length: ROWS }, () => -50);
+const upper = Float64Array.from({ length: ROWS }, () => 50);
+const s = new Series(data);
+const lowerArr = Array.from(lower);
+const upperArr = Array.from(upper);
+
+const dfCols: Record<string, number[]> = {};
+for (let c = 0; c < 5; c++) {
+  dfCols[`col${c}`] = Array.from({ length: ROWS }, (_, i) => Math.sin((i + c) * 0.01) * 200);
+}
+const df = new DataFrame(dfCols);
+
+for (let i = 0; i < WARMUP; i++) {
+  clipAdvancedSeries(s, { lower: lowerArr, upper: upperArr });
+  clipAdvancedDataFrame(df, { lower: -50, upper: 50 });
+}
+
+const start = performance.now();
+for (let i = 0; i < ITERATIONS; i++) {
+  clipAdvancedSeries(s, { lower: lowerArr, upper: upperArr });
+  clipAdvancedDataFrame(df, { lower: -50, upper: 50 });
+}
+const total = performance.now() - start;
+
+console.log(JSON.stringify({ function: "clip_advanced", mean_ms: total / ITERATIONS, iterations: ITERATIONS, total_ms: total }));

benchmarks/tsb/bench_dataframe_sem_var.ts

@@ -0,0 +1,30 @@
+/**
+ * Benchmark: varDataFrame / semDataFrame — variance and SEM on a 10k×10 DataFrame.
+ * Outputs JSON: {"function": "dataframe_sem_var", "mean_ms": ..., "iterations": ..., "total_ms": ...}
+ */
+import { DataFrame, varDataFrame, semDataFrame } from "../../src/index.ts";
+
+const ROWS = 10_000;
+const COLS = 10;
+const WARMUP = 5;
+const ITERATIONS = 20;
+
+const columns: Record<string, number[]> = {};
+for (let c = 0; c < COLS; c++) {
+  columns[`col${c}`] = Array.from({ length: ROWS }, (_, i) => Math.sin((i + c) * 0.01) * 100);
+}
+const df = new DataFrame(columns);
+
+for (let i = 0; i < WARMUP; i++) {
+  varDataFrame(df);
+  semDataFrame(df);
+}
+
+const start = performance.now();
+for (let i = 0; i < ITERATIONS; i++) {
+  varDataFrame(df);
+  semDataFrame(df);
+}
+const total = performance.now() - start;
+
+console.log(JSON.stringify({ function: "dataframe_sem_var", mean_ms: total / ITERATIONS, iterations: ITERATIONS, total_ms: total }));