added performance plotting

Author: flexatone

doc/articles/first_true_2d.py

@@ -0,0 +1,362 @@
+
+
+
+import os
+import sys
+import timeit
+import typing as tp
+from itertools import repeat
+
+from arraykit import first_true_2d as ak_first_true_2d
+from arrayredox import first_true_2d as ar_first_true_2d
+import arraykit as ak
+
+import matplotlib.pyplot as plt
+import numpy as np
+import pandas as pd
+
+sys.path.append(os.getcwd())
+
+
+
+class ArrayProcessor:
+    NAME = ''
+    SORT = -1
+
+    def __init__(self, array: np.ndarray):
+        self.array = array
+
+#-------------------------------------------------------------------------------
+class AKFirstTrueAxis0Forward(ArrayProcessor):
+    NAME = 'ak.first_true_2d(forward=True, axis=0)'
+    SORT = 0
+
+    def __call__(self):
+        _ = ak_first_true_2d(self.array, forward=True, axis=0)
+
+class AKFirstTrueAxis1Forward(ArrayProcessor):
+    NAME = 'ak.first_true_2d(forward=True, axis=1)'
+    SORT = 1
+
+    def __call__(self):
+        _ = ak_first_true_2d(self.array, forward=True, axis=1)
+
+class AKFirstTrueAxis0Reverse(ArrayProcessor):
+    NAME = 'ak.first_true_2d(forward=False, axis=0)'
+    SORT = 2
+
+    def __call__(self):
+        _ = ak_first_true_2d(self.array, forward=False, axis=0)
+
+class AKFirstTrueAxis1Reverse(ArrayProcessor):
+    NAME = 'ak.first_true_2d(forward=False, axis=1)'
+    SORT = 3
+
+    def __call__(self):
+        _ = ak_first_true_2d(self.array, forward=False, axis=1)
+
+
+
+#-------------------------------------------------------------------------------
+class ARFirstTrueAxis0Forward(ArrayProcessor):
+    NAME = 'ar.first_true_2d(forward=True, axis=0)'
+    SORT = 10
+
+    def __call__(self):
+        _ = ar_first_true_2d(self.array, forward=True, axis=0)
+
+class ARFirstTrueAxis1Forward(ArrayProcessor):
+    NAME = 'ar.first_true_2d(forward=True, axis=1)'
+    SORT = 11
+
+    def __call__(self):
+        _ = ar_first_true_2d(self.array, forward=True, axis=1)
+
+class ARFirstTrueAxis0Reverse(ArrayProcessor):
+    NAME = 'ar.first_true_2d(forward=False, axis=0)'
+    SORT = 12
+
+    def __call__(self):
+        _ = ar_first_true_2d(self.array, forward=False, axis=0)
+
+class ARFirstTrueAxis1Reverse(ArrayProcessor):
+    NAME = 'ar.first_true_2d(forward=False, axis=1)'
+    SORT = 13
+
+    def __call__(self):
+        _ = ar_first_true_2d(self.array, forward=False, axis=1)
+
+
+#-------------------------------------------------------------------------------
+
+
+class NPNonZero(ArrayProcessor):
+    NAME = 'np.nonzero()'
+    SORT = 3
+
+    def __call__(self):
+        x, y = np.nonzero(self.array)
+        # list(zip(x, y)) # simulate iteration
+
+
+class NPArgMaxAxis0(ArrayProcessor):
+    NAME = 'np.any(axis=0), np.argmax(axis=0)'
+    SORT = 4
+
+    def __call__(self):
+        _ = ~np.any(self.array, axis=0)
+        _ = np.argmax(self.array, axis=0)
+
+class NPArgMaxAxis1(ArrayProcessor):
+    NAME = 'np.any(axis=1), np.argmax(axis=1)'
+    SORT = 4
+
+    def __call__(self):
+        _ = ~np.any(self.array, axis=1)
+        _ = np.argmax(self.array, axis=1)
+
+
+
+#-------------------------------------------------------------------------------
+NUMBER = 100
+
+def seconds_to_display(seconds: float) -> str:
+    seconds /= NUMBER
+    if seconds < 1e-4:
+        return f'{seconds * 1e6: .1f} (µs)'
+    if seconds < 1e-1:
+        return f'{seconds * 1e3: .1f} (ms)'
+    return f'{seconds: .1f} (s)'
+
+
+def plot_performance(frame):
+    fixture_total = len(frame['fixture'].unique())
+    cat_total = len(frame['size'].unique())
+    processor_total = len(frame['cls_processor'].unique())
+    fig, axes = plt.subplots(cat_total, fixture_total)
+
+    # cmap = plt.get_cmap('terrain')
+    cmap = plt.get_cmap('plasma')
+
+    color = cmap(np.arange(processor_total) / processor_total)
+
+    # category is the size of the array
+    for cat_count, (cat_label, cat) in enumerate(frame.groupby('size')):
+        for fixture_count, (fixture_label, fixture) in enumerate(
+                cat.groupby('fixture')):
+            ax = axes[cat_count][fixture_count]
+
+            # set order
+            fixture['sort'] = [f.SORT for f in fixture['cls_processor']]
+            fixture = fixture.sort_values('sort')
+
+            results = fixture['time'].values.tolist()
+            names = [cls.NAME for cls in fixture['cls_processor']]
+            # x = np.arange(len(results))
+            names_display = names
+            post = ax.bar(names_display, results, color=color)
+
+            density, position = fixture_label.split('-')
+            # cat_label is the size of the array
+            title = f'{cat_label:.0e}\n{FixtureFactory.DENSITY_TO_DISPLAY[density]}\n{FixtureFactory.POSITION_TO_DISPLAY[position]}'
+
+            ax.set_title(title, fontsize=6)
+            ax.set_box_aspect(0.75) # makes taller tan wide
+            time_max = fixture['time'].max()
+            ax.set_yticks([0, time_max * 0.5, time_max])
+            ax.set_yticklabels(['',
+                    seconds_to_display(time_max * .5),
+                    seconds_to_display(time_max),
+                    ], fontsize=6)
+            # ax.set_xticks(x, names_display, rotation='vertical')
+            ax.tick_params(
+                    axis='x',
+                    which='both',
+                    bottom=False,
+                    top=False,
+                    labelbottom=False,
+                    )
+
+    fig.set_size_inches(9, 3.5) # width, height
+    fig.legend(post, names_display, loc='center right', fontsize=6)
+    # horizontal, vertical
+    fig.text(.05, .96, f'ak_first_true_2d() Performance: {NUMBER} Iterations', fontsize=10)
+    fig.text(.05, .90, get_versions(), fontsize=6)
+
+    fp = '/tmp/first_true.png'
+    plt.subplots_adjust(
+            left=0.075,
+            bottom=0.05,
+            right=0.75,
+            top=0.85,
+            wspace=1, # width
+            hspace=0.1,
+            )
+    # plt.rcParams.update({'font.size': 22})
+    plt.savefig(fp, dpi=300)
+
+    if sys.platform.startswith('linux'):
+        os.system(f'eog {fp}&')
+    else:
+        os.system(f'open {fp}')
+
+
+#-------------------------------------------------------------------------------
+
+class FixtureFactory:
+    NAME = ''
+
+    @staticmethod
+    def get_array(size: int) -> np.ndarray:
+        return np.full(size, False, dtype=bool)
+
+    def _get_array_filled(
+            size: int,
+            start_third: int, # 1 or 2
+            density: float, # less than 1
+            ) -> np.ndarray:
+        a = FixtureFactory.get_array(size)
+        count = size * density
+        start = int(len(a) * (start_third/3))
+        length = len(a) - start
+        step = int(length / count)
+        fill = np.arange(start, len(a), step)
+        a[fill] = True
+        return a
+
+    @classmethod
+    def get_label_array(cls, size: int) -> tp.Tuple[str, np.ndarray]:
+        array = cls.get_array(size)
+        return cls.NAME, array
+
+    DENSITY_TO_DISPLAY = {
+        'single': '1 True',
+        'tenth': '10% True',
+        'third': '33% True',
+    }
+
+    POSITION_TO_DISPLAY = {
+        'first_third': 'Fill 1/3 to End',
+        'second_third': 'Fill 2/3 to End',
+    }
+
+
+class FFSingleFirstThird(FixtureFactory):
+    NAME = 'single-first_third'
+
+    @staticmethod
+    def get_array(size: int) -> np.ndarray:
+        a = FixtureFactory.get_array(size)
+        a[int(len(a) * (1/3))] = True
+        return a
+
+class FFSingleSecondThird(FixtureFactory):
+    NAME = 'single-second_third'
+
+    @staticmethod
+    def get_array(size: int) -> np.ndarray:
+        a = FixtureFactory.get_array(size)
+        a[int(len(a) * (2/3))] = True
+        return a
+
+
+class FFTenthPostFirstThird(FixtureFactory):
+    NAME = 'tenth-first_third'
+
+    @classmethod
+    def get_array(cls, size: int) -> np.ndarray:
+        return cls._get_array_filled(size, start_third=1, density=.1)
+
+
+class FFTenthPostSecondThird(FixtureFactory):
+    NAME = 'tenth-second_third'
+
+    @classmethod
+    def get_array(cls, size: int) -> np.ndarray:
+        return cls._get_array_filled(size, start_third=2, density=.1)
+
+
+class FFThirdPostFirstThird(FixtureFactory):
+    NAME = 'third-first_third'
+
+    @classmethod
+    def get_array(cls, size: int) -> np.ndarray:
+        return cls._get_array_filled(size, start_third=1, density=1/3)
+
+
+class FFThirdPostSecondThird(FixtureFactory):
+    NAME = 'third-second_third'
+
+    @classmethod
+    def get_array(cls, size: int) -> np.ndarray:
+        return cls._get_array_filled(size, start_third=2, density=1/3)
+
+
+def get_versions() -> str:
+    import platform
+    return f'OS: {platform.system()} / ArrayKit: {ak.__version__} / NumPy: {np.__version__}\n'
+
+
+CLS_PROCESSOR = (
+    AKFirstTrueAxis0Forward,
+    AKFirstTrueAxis1Forward,
+    AKFirstTrueAxis0Reverse,
+    AKFirstTrueAxis1Reverse,
+
+    ARFirstTrueAxis0Forward,
+    ARFirstTrueAxis1Forward,
+    ARFirstTrueAxis0Reverse,
+    ARFirstTrueAxis1Reverse,
+
+    # NPNonZero,
+    # NPArgMaxAxis0,
+    # NPArgMaxAxis1
+    )
+
+CLS_FF = (
+    FFSingleFirstThird,
+    FFSingleSecondThird,
+    FFTenthPostFirstThird,
+    FFTenthPostSecondThird,
+    FFThirdPostFirstThird,
+    FFThirdPostSecondThird,
+)
+
+
+def run_test():
+    records = []
+    for size in (100_000, 1_000_000, 10_000_000):
+        for ff in CLS_FF:
+            fixture_label, fixture = ff.get_label_array(size)
+            # TEMP
+            fixture = fixture.reshape(size // 10, 10)
+
+            for cls in CLS_PROCESSOR:
+                runner = cls(fixture)
+
+                record = [cls, NUMBER, fixture_label, size]
+                print(record)
+                try:
+                    result = timeit.timeit(
+                            f'runner()',
+                            globals=locals(),
+                            number=NUMBER)
+                except OSError:
+                    result = np.nan
+                finally:
+                    pass
+                record.append(result)
+                records.append(record)
+
+    f = pd.DataFrame.from_records(records,
+            columns=('cls_processor', 'number', 'fixture', 'size', 'time')
+            )
+    print(f)
+    plot_performance(f)
+
+if __name__ == '__main__':
+
+    run_test()
+
+
+

src/lib.rs

@@ -400,6 +400,7 @@ pub fn first_true_2d<'py>(
     let prepped = prepare_array_for_axis(py, array, axis)?;
     let view = unsafe { prepped.as_array() };
 
+    // let view = array.as_array();
     // NOTE: these are rows in the view, not always the same as rows
     let rows = view.nrows();
     let mut result = Vec::with_capacity(rows);
@@ -408,7 +409,6 @@ pub fn first_true_2d<'py>(
         const LANES: usize = 32;
         let ones = u8x32::splat(1);
 
-
         for row in 0..rows {
             let mut found = -1;
             let row_slice = &view.row(row);