Added pre-commit config with black, isort, and pyupgrade

.pre-commit-config.yaml

@@ -0,0 +1,29 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.4.0
+    hooks:
+      - id: check-added-large-files
+      - id: check-case-conflict
+      - id: check-yaml
+      - id: debug-statements
+      - id: end-of-file-fixer
+      - id: trailing-whitespace
+
+  - repo: https://github.com/pycqa/isort
+    rev: 5.12.0
+    hooks:
+      - id: isort
+        name: isort (python)
+        args:
+          - "--filter-files"
+
+  - repo: https://github.com/asottile/pyupgrade
+    rev: v3.10.1
+    hooks:
+      - id: pyupgrade
+        args: ["--py38-plus"]
+
+  - repo: https://github.com/psf/black
+    rev: 23.9.1
+    hooks:
+      - id: black

comparison/benchmark.py

@@ -1,7 +1,8 @@
 # Script to compare the speedup provided by fast-histogram
 
+from timeit import repeat, timeit
+
 import numpy as np
-from timeit import timeit, repeat
 
 SETUP_1D = """
 import numpy as np
@@ -29,42 +30,75 @@
 
 
 def time_stats(stmt=None, setup=None):
-
     # Call once to check how long it takes
     time_single = timeit(stmt=stmt, setup=setup, number=1)
 
     # Find out how many times we can call it. We always call it at least three
     # times for accuracy
     number = max(3, int(TARGET_TIME / time_single))
 
-    print(' -> estimated time to complete test: {0:.1f}s'.format(time_single * 10 * number))
+    print(f" -> estimated time to complete test: {time_single * 10 * number:.1f}s")
 
     times = repeat(stmt=stmt, setup=setup, repeat=10, number=number)
 
     return np.min(times) / number, np.mean(times) / number, np.median(times) / number
 
 
-FMT_HEADER = '# {:7s}' + ' {:10s}' * 12 + '\n'
-FMT = '{:9d}' + ' {:10.7e}' * 12 + '\n'
-
-with open('benchmark_times.txt', 'w') as f:
-
-    f.write(FMT_HEADER.format('size',
-                              'np_1d_min', 'np_1d_mean', 'np_1d_median', 'fa_1d_min', 'fa_1d_mean', 'fa_1d_median',
-                              'np_2d_min', 'np_2d_mean', 'np_2d_median', 'fa_2d_min', 'fa_2d_mean', 'fa_2d_median'))
+FMT_HEADER = "# {:7s}" + " {:10s}" * 12 + "\n"
+FMT = "{:9d}" + " {:10.7e}" * 12 + "\n"
+
+with open("benchmark_times.txt", "w") as f:
+    f.write(
+        FMT_HEADER.format(
+            "size",
+            "np_1d_min",
+            "np_1d_mean",
+            "np_1d_median",
+            "fa_1d_min",
+            "fa_1d_mean",
+            "fa_1d_median",
+            "np_2d_min",
+            "np_2d_mean",
+            "np_2d_median",
+            "fa_2d_min",
+            "fa_2d_mean",
+            "fa_2d_median",
+        )
+    )
 
     for log10_size in range(0, 9):
-
-        size = int(10 ** log10_size)
-
-        print('Running benchmarks for size={0}'.format(size))
-
-        np_1d_min, np_1d_mean, np_1d_median = time_stats(stmt=NUMPY_1D_STMT, setup=SETUP_1D.format(size=size))
-        fa_1d_min, fa_1d_mean, fa_1d_median = time_stats(stmt=FAST_1D_STMT, setup=SETUP_1D.format(size=size))
-        np_2d_min, np_2d_mean, np_2d_median = time_stats(stmt=NUMPY_2D_STMT, setup=SETUP_2D.format(size=size))
-        fa_2d_min, fa_2d_mean, fa_2d_median = time_stats(stmt=FAST_2D_STMT, setup=SETUP_2D.format(size=size))
-
-        f.write(FMT.format(size,
-                           np_1d_min, np_1d_mean, np_1d_median, fa_1d_min, fa_1d_mean, fa_1d_median,
-                           np_2d_min, np_2d_mean, np_2d_median, fa_2d_min, fa_2d_mean, fa_2d_median))
+        size = int(10**log10_size)
+
+        print(f"Running benchmarks for size={size}")
+
+        np_1d_min, np_1d_mean, np_1d_median = time_stats(
+            stmt=NUMPY_1D_STMT, setup=SETUP_1D.format(size=size)
+        )
+        fa_1d_min, fa_1d_mean, fa_1d_median = time_stats(
+            stmt=FAST_1D_STMT, setup=SETUP_1D.format(size=size)
+        )
+        np_2d_min, np_2d_mean, np_2d_median = time_stats(
+            stmt=NUMPY_2D_STMT, setup=SETUP_2D.format(size=size)
+        )
+        fa_2d_min, fa_2d_mean, fa_2d_median = time_stats(
+            stmt=FAST_2D_STMT, setup=SETUP_2D.format(size=size)
+        )
+
+        f.write(
+            FMT.format(
+                size,
+                np_1d_min,
+                np_1d_mean,
+                np_1d_median,
+                fa_1d_min,
+                fa_1d_mean,
+                fa_1d_median,
+                np_2d_min,
+                np_2d_mean,
+                np_2d_median,
+                fa_2d_min,
+                fa_2d_mean,
+                fa_2d_median,
+            )
+        )
         f.flush()

comparison/plot.py

@@ -1,21 +1,33 @@
 # Script to make the comparison plot for the benchmark
 
-import numpy as np
 import matplotlib.pyplot as plt
+import numpy as np
 
-(size,
- np_1d_min, np_1d_mean, np_1d_median, fa_1d_min, fa_1d_mean, fa_1d_median,
- np_2d_min, np_2d_mean, np_2d_median, fa_2d_min, fa_2d_mean, fa_2d_median) = np.loadtxt('benchmark_times.txt', unpack=True)
+(
+    size,
+    np_1d_min,
+    np_1d_mean,
+    np_1d_median,
+    fa_1d_min,
+    fa_1d_mean,
+    fa_1d_median,
+    np_2d_min,
+    np_2d_mean,
+    np_2d_median,
+    fa_2d_min,
+    fa_2d_mean,
+    fa_2d_median,
+) = np.loadtxt("benchmark_times.txt", unpack=True)
 
 fig = plt.figure()
 ax = fig.add_subplot(1, 1, 1)
-ax.plot(size, np_1d_min / fa_1d_min, color=(34 / 255, 122 / 255, 181 / 255), label='1D')
-ax.plot(size, np_2d_min / fa_2d_min, color=(255 / 255, 133 / 255, 25 / 255), label='2D')
-ax.set_xscale('log')
+ax.plot(size, np_1d_min / fa_1d_min, color=(34 / 255, 122 / 255, 181 / 255), label="1D")
+ax.plot(size, np_2d_min / fa_2d_min, color=(255 / 255, 133 / 255, 25 / 255), label="2D")
+ax.set_xscale("log")
 ax.set_xlim(0.3, 3e8)
 ax.set_ylim(1, 35)
 ax.grid()
-ax.set_xlabel('Array size')
-ax.set_ylabel('Speedup (fast-histogram / numpy)')
+ax.set_xlabel("Array size")
+ax.set_ylabel("Speedup (fast-histogram / numpy)")
 ax.legend()
-fig.savefig('speedup_compared.png', bbox_inches='tight')
+fig.savefig("speedup_compared.png", bbox_inches="tight")

fast_histogram/histogram.py

@@ -1,19 +1,19 @@
-from __future__ import division
-
 import numbers
 
 import numpy as np
 
-from ._histogram_core import (_histogram1d,
-                              _histogram2d,
-                              _histogramdd,
-                              _histogram1d_weighted,
-                              _histogram2d_weighted,
-                              _histogramdd_weighted)
+from ._histogram_core import (
+    _histogram1d,
+    _histogram1d_weighted,
+    _histogram2d,
+    _histogram2d_weighted,
+    _histogramdd,
+    _histogramdd_weighted,
+)
 
-NUMERICAL_TYPES = {'f', 'i', 'u'}
+NUMERICAL_TYPES = {"f", "i", "u"}
 
-__all__ = ['histogram1d', 'histogram2d', 'histogramdd']
+__all__ = ["histogram1d", "histogram2d", "histogramdd"]
 
 
 def histogram1d(x, bins, range, weights=None):
@@ -40,7 +40,7 @@ def histogram1d(x, bins, range, weights=None):
     nx = bins
 
     if not np.isscalar(bins):
-        raise TypeError('bins should be an integer')
+        raise TypeError("bins should be an integer")
 
     xmin, xmax = range
 
@@ -66,7 +66,9 @@ def histogram1d(x, bins, range, weights=None):
     else:
         weights = np.atleast_1d(weights)
         if weights.dtype.kind not in NUMERICAL_TYPES:
-            raise TypeError("weights is not or cannot be converted to a numerical array")
+            raise TypeError(
+                "weights is not or cannot be converted to a numerical array"
+            )
         return _histogram1d_weighted(x, weights, nx, xmin, xmax)
 
 
@@ -99,7 +101,7 @@ def histogram2d(x, y, bins, range, weights=None):
         nx, ny = bins
 
     if not np.isscalar(nx) or not np.isscalar(ny):
-        raise TypeError('bins should be an iterable of two integers')
+        raise TypeError("bins should be an iterable of two integers")
 
     (xmin, xmax), (ymin, ymax) = range
 
@@ -141,7 +143,9 @@ def histogram2d(x, y, bins, range, weights=None):
     else:
         weights = np.atleast_1d(weights)
         if weights.dtype.kind not in NUMERICAL_TYPES:
-            raise TypeError("weights is not or cannot be converted to a numerical array")
+            raise TypeError(
+                "weights is not or cannot be converted to a numerical array"
+            )
         return _histogram2d_weighted(x, y, weights, nx, xmin, xmax, ny, ymin, ymax)
 
 
@@ -206,7 +210,9 @@ def histogramdd(sample, bins, range, weights=None):
         raise ValueError("number of ranges does not equal number of dimensions")
     for i, r in enumerate(range):
         if not len(r) == 2:
-            raise ValueError("should pass a minimum and maximum value for each dimension")
+            raise ValueError(
+                "should pass a minimum and maximum value for each dimension"
+            )
         if r[0] >= r[1]:
             raise ValueError("each range should be strictly increasing")
         _range[i][0] = r[0]
@@ -217,5 +223,7 @@ def histogramdd(sample, bins, range, weights=None):
     else:
         weights = np.atleast_1d(weights)
         if weights.dtype.kind not in NUMERICAL_TYPES:
-            raise TypeError("weights is not or cannot be converted to a numerical array")
+            raise TypeError(
+                "weights is not or cannot be converted to a numerical array"
+            )
         return _histogramdd_weighted(_sample, _bins, _range, weights)