Add: implement benchmark tests for curve rendering with large datasets

Author: PierreRaybaut

plotpy/builder/curvemarker.py

@@ -77,7 +77,7 @@ def __set_baseparam(
         param: CurveParam | MarkerParam,
         color: str | None = None,
         linestyle: str | None = None,
-        linewidth: int | None = None,
+        linewidth: float | None = None,
         marker: str | None = None,
         markersize: int | None = None,
         markerfacecolor: str | None = None,
@@ -116,7 +116,7 @@ def __set_param(
         title: str | None = None,
         color: str | None = None,
         linestyle: str | None = None,
-        linewidth: int | None = None,
+        linewidth: float | None = None,
         marker: str | None = None,
         markersize: int | None = None,
         markerfacecolor: str | None = None,
@@ -313,7 +313,7 @@ def curve(
         title: str = "",
         color: str | None = None,
         linestyle: str | None = None,
-        linewidth: int | None = None,
+        linewidth: float | None = None,
         marker: str | None = None,
         markersize: int | None = None,
         markerfacecolor: str | None = None,

plotpy/tests/benchmarks/test_bigcurves.py

@@ -0,0 +1,89 @@
+# -*- coding: utf-8 -*-
+#
+# Licensed under the terms of the BSD 3-Clause
+# (see plotpy/LICENSE for details)
+
+"""Curve plotting test with large datasets"""
+
+# guitest: show
+
+import timeit
+
+import numpy as np
+import pytest
+from guidata.qthelpers import qt_app_context
+from qtpy import QtWidgets as QW
+
+from plotpy.builder import make
+from plotpy.tests import vistools as ptv
+
+
+def __create_curve(npts: int, linewidth: float):
+    """Create a curve with random data"""
+    x = np.linspace(-10, 10, npts)
+    y = np.random.normal(size=npts)
+    curve = make.curve(x, y, color="g", linewidth=linewidth)
+    return curve
+
+
+def __show_curve(curve, benchmark: bool = False):
+    """Show the curve in a plot window"""
+    with qt_app_context(exec_loop=not benchmark):
+        _win = ptv.show_items(
+            [curve],
+            wintitle=test_curve_rendering_performance.__doc__,
+            title="Curves",
+            plot_type="curve",
+            disable_readonly_for_items=False,
+        )
+        if benchmark:
+            QW.QApplication.processEvents()
+    return _win
+
+
+@pytest.mark.skip(reason="Not relevant in automated test suite")
+def test_interactive_curve_rendering(npts: int = 500000, linewidth: float = 1.0):
+    """Interactive test for curve rendering with large datasets"""
+    curve = __create_curve(npts, linewidth)
+    _win = __show_curve(curve, benchmark=False)
+
+
+@pytest.mark.skip(reason="Not relevant in automated test suite")
+def test_curve_rendering_performance(npts: int = 500000, linewidth: float = 1.0):
+    """Test curve rendering performance with large datasets"""
+    curve = __create_curve(npts, linewidth)
+
+    # Show once before benchmarking
+    _win = __show_curve(curve, benchmark=True)
+
+    # Benchmark with multiple runs
+    n_runs = 5
+    times = []
+
+    print("Benchmarking in progress", end="")
+    for i in range(n_runs):
+        start_time = timeit.default_timer()
+        _win = __show_curve(curve, benchmark=True)
+        elapsed_time = timeit.default_timer() - start_time
+        times.append(elapsed_time)
+        print(".", end="", flush=True)
+    print(" done.")
+
+    # Display statistics
+    print(
+        f"Results ({n_runs} runs): {np.mean(times):.4f} ± {np.std(times):.4f} seconds"
+    )
+
+
+if __name__ == "__main__":
+    # test_interactive_curve_rendering(10000, 2.0, False)
+    # Test with different linewidths and antialiasing settings
+    print("\n" + "=" * 70)
+    print("Curve Rendering Performance Benchmark")
+    print("=" * 70)
+
+    print("\n--- Test 1: linewidth=1.0 ---")
+    test_curve_rendering_performance(10000, 1.0)
+
+    print("\n--- Test 2: linewidth=2.0 ---")
+    test_curve_rendering_performance(10000, 2.0)