Metrics modules

setup.py

@@ -1,12 +1,16 @@
 from distutils.core import setup
 
+
 setup(
     name='poprank',
-    version='0.1',
+    version='1.0.0',
     description='Rating Mechanisms',
     author='Manfred Diaz, Aurelien Buck-Kaeffer',
     author_email='',
     url='https://www.python.org/sigs/distutils-sig/',
     package_dir={'': 'src'},
     packages=['poprank'],
+    install_requires=[
+        'popcore @ git+ssh://git@github.com/poprl/popcore@v1.0.0-beta#egg=popcore',
+    ]
 )

src/poprank/metrics/_base.py

@@ -0,0 +1,38 @@
+from abc import ABC, abstractmethod
+from typing import Any
+
+import numpy as np
+
+
+class Metric(ABC):
+
+    def __init__(self, name: str) -> None:
+        super().__init__()
+        self.name = name
+
+    @abstractmethod
+    def _compute(
+        self, x: np.ndarray, y: np.ndarray, *args, **kwargs
+    ) -> float:
+        raise NotImplementedError()
+
+    @abstractmethod
+    def _max(self, n: int) -> int:
+        raise NotImplementedError()
+
+    def __call__(
+        self, x: np.ndarray, y: np.ndarray, *args: Any, **kwds: Any
+    ) -> float:
+        return self._compute(x, y, *args, **kwds)
+
+    def max(self, n: int) -> int:
+        """
+            Computes the upper bound value of the metric
+            on the space of n-ranked permutations.
+
+        :param n: number of alternatives.
+        :type n: int
+        :return: metric upper bound.
+        :rtype: int
+        """
+        return self._max(n)

src/poprank/metrics/hamming.py

@@ -0,0 +1,16 @@
+import numpy as np
+from ..functional.metrics import hamming
+from ._base import Metric
+
+
+class Hamming(Metric):
+    def __init__(self) -> None:
+        super().__init__("hamming")
+
+    def _compute(
+        self, x: np.ndarray, y: np.ndarray, *args, **kwargs
+    ) -> float:
+        return hamming(x, y, *args, **kwargs)
+
+    def _max(self, n: int) -> int:
+        return n

src/poprank/metrics/kendall.py

@@ -0,0 +1,18 @@
+import numpy as np
+
+from ..functional.metrics import kendall
+from ._base import Metric
+
+
+class Kendall(Metric):
+
+    def __init__(self) -> None:
+        super().__init__("kendall")
+
+    def _compute(
+        self, x: np.ndarray, y: np.ndarray, *args, **kwargs
+    ) -> float:
+        return kendall(x, y, *args, **kwargs)
+
+    def _max(self, n: int) -> int:
+        return n * (n - 1) / 2