Drop Python 3.9 and Numpy 1.23 support.

Following release of Numpy 2.1.0.

.readthedocs.yaml

@@ -8,6 +8,6 @@ python:
 build:
   os: ubuntu-22.04
   tools:
-    python: "3.9"
+    python: "3.10"
 sphinx:
    configuration: docs/conf.py

.travis.yml

@@ -1,17 +1,15 @@
 dist: jammy
 language: python
 jobs:
-  - python: 3.9
-    env:
-      - MPYC_NOPRSS=1
-    install: pip install numpy==1.23.* gmpy2 uvloop
-  - python: pypy3.9-7.3.9
   - python: 3.10
     env:
       - MPYC_NONUMPY=1
       - MPYC_NOGMPY=1
       - MPYC_NOUVLOOP=1
+  - python: pypy3.10-7.3.16
   - python: 3.11
+    env:
+      - MPYC_NOPRSS=1
     install: pip install numpy==1.24.* gmpy2
   - python: 3.12
     install: pip install numpy uvloop

README.md

@@ -25,7 +25,7 @@ The [MPyC homepage](https://www.win.tue.nl/~berry/mpyc/) has some more info and
 
 ## Installation
 
-Pure Python, no dependencies. Python 3.9+ (following [SPEC 0 -- Minimum Supported Dependencies](https://scientific-python.org/specs/spec-0000/)).
+Pure Python, no dependencies. Python 3.10+ (following [SPEC 0 -- Minimum Supported Dependencies](https://scientific-python.org/specs/spec-0000/)).
 
 Run `pip install .` in the root directory (containing file `pyproject.toml`).\
 Or, run `pip install -e .`, if you want to edit the MPyC source files.

docs/conf.py

@@ -14,6 +14,7 @@
 #
 import os
 import sys
+from mpyc.__init__ import __version__
 # sys.path.insert(0, os.path.abspath('.'))
 sys.path.insert(0, os.path.abspath('..'))
 
@@ -23,7 +24,6 @@
 copyright = '2018 - 2024, Berry Schoenmakers'
 author = 'Berry Schoenmakers'
 
-from mpyc.__init__ import __version__
 # The short X.Y version
 version = __version__
 # The full version, including alpha/beta/rc tags

docs/install.rst

@@ -1,7 +1,7 @@
 MPyC installation
 =================
 
-MPyC runs on Windows/Mac/Linux platforms supporting Python 3.9+.
+MPyC runs on Windows/Mac/Linux platforms supporting Python 3.10+.
 There are no hard dependencies on other Python packages.
 
 If you first want to try MPyC online, click
@@ -59,7 +59,7 @@ See the ``README.md`` files in these directories for more information.
 Deployment
 ----------
 
-Currently Python 3.9+ is the only requirement.
+Currently Python 3.10+ is the only requirement.
 
 Written in pure Python, without any dependencies, MPyC should run on common
 Windows/Mac/Linux platforms, including the 32-bit Raspberry Pi, if you like.

mpyc/__init__.py

@@ -29,7 +29,7 @@
 and statistics (securely mimicking Python’s statistics module).
 """
 
-__version__ = '0.10.2'
+__version__ = '0.10.3'
 
 import os
 import sys

mpyc/fingroups.py

@@ -1056,14 +1056,15 @@ def _EllipticCurve(curvename, coordinates):
             raise ValueError('invalid curvename')
 
         name = f'E({gf.__name__}){curvename}{coordinates}'
-        if coordinates == 'extended':
-            base = EdwardsExtended
-        elif coordinates == 'affine':
-            base = EdwardsAffine
-        elif coordinates == 'projective':
-            base = EdwardsProjective
-        else:
-            raise ValueError('invalid coordinates')
+        match coordinates:
+            case 'extended':
+                base = EdwardsExtended
+            case 'affine':
+                base = EdwardsAffine
+            case 'projective':
+                base = EdwardsProjective
+            case _:
+                raise ValueError('invalid coordinates')
 
         EC = type(name, (base,), {'__slots__': ()})
         EC.field = gf
@@ -1097,14 +1098,15 @@ def _EllipticCurve(curvename, coordinates):
             raise ValueError('invalid curvename')
 
         name = f'E({gf.__name__}){curvename}{coordinates}'
-        if coordinates == 'jacobian':
-            base = WeierstrassJacobian
-        elif coordinates == 'affine':
-            base = WeierstrassAffine
-        elif coordinates == 'projective':
-            base = WeierstrassProjective
-        else:
-            raise ValueError('invalid coordinates')
+        match coordinates:
+            case 'jacobian':
+                base = WeierstrassJacobian
+            case 'affine':
+                base = WeierstrassAffine
+            case 'projective':
+                base = WeierstrassProjective
+            case _:
+                raise ValueError('invalid coordinates')
 
         EC = type(name, (base,), {'__slots__': ()})
         EC.field = gf
@@ -1128,14 +1130,15 @@ def _EllipticCurve(curvename, coordinates):
         gf = GF(p)
 
         name = f'E({gf.__name__}){curvename}{coordinates}'
-        if coordinates == 'jacobian':
-            base = WeierstrassJacobian
-        elif coordinates == 'affine':
-            base = WeierstrassAffine
-        elif coordinates == 'projective':
-            base = WeierstrassProjective
-        else:
-            raise ValueError('invalid coordinates')
+        match coordinates:
+            case 'jacobian':
+                base = WeierstrassJacobian
+            case 'affine':
+                base = WeierstrassAffine
+            case 'projective':
+                base = WeierstrassProjective
+            case _:
+                raise ValueError('invalid coordinates')
 
         EC = type(name, (base,), {'__slots__': ()})
         EC.field = gf

mpyc/mpctools.py

@@ -4,7 +4,7 @@
 The alternative implementations can be used as drop-in replacements, however,
 potentially enhancing the performance when used in secure computations. More
 specifically, these implementations are aimed at reducing the overall round
-complexity, possible at the expense of increasing overall space complexity,
+complexity, possibly at the expense of increasing overall space complexity,
 time complexity, and communication complexity.
 """
 

mpyc/numpy.py

@@ -161,8 +161,8 @@ def _item_shape(shape, key):
     np._matmul_shape = _matmul_shape
     np._item_shape = _item_shape
 
-    if np.lib.NumpyVersion(np.__version__) < '1.23.0':
-        logging.warning(f'NumPy {np.__version__} not (fully) supported. Upgrade to NumPy 1.23+.')
+    if np.lib.NumpyVersion(np.__version__) < '1.24.0':
+        logging.warning(f'NumPy {np.__version__} not (fully) supported. Upgrade to NumPy 1.24+.')
 except ImportError:
     del _matmul_shape
     del _item_shape

mpyc/statistics.py

@@ -22,7 +22,6 @@
 function in Python's statistics module.
 """
 
-import sys
 from math import fsum, sqrt
 import statistics
 from mpyc import asyncoro
@@ -506,14 +505,7 @@ def covariance(x, y):
 
     sectype = type(x[0])  # all elts of x assumed of same type
     if not issubclass(sectype, SecureObject):
-        if sys.version_info.minor >= 10:
-            return statistics.covariance(x, y)
-
-        # inline code of statistics.covariance() copied from Python 3.10.0:
-        xbar = fsum(x) / n
-        ybar = fsum(y) / n
-        sxy = fsum((xi - xbar) * (yi - ybar) for xi, yi in zip(x, y))
-        return sxy / (n - 1)
+        return statistics.covariance(x, y)
 
     if issubclass(sectype, SecureFixedPoint):
         xbar = runtime.sum(x) / n
@@ -549,20 +541,7 @@ def correlation(x, y):
 
     sectype = type(x[0])  # all elts of x assumed of same type
     if not issubclass(sectype, SecureObject):
-        if sys.version_info.minor >= 10:
-            return statistics.correlation(x, y)
-
-        # inline code of statistics.correlation() copied from Python 3.10.0:
-        xbar = fsum(x) / n
-        ybar = fsum(y) / n
-        sxy = fsum((xi - xbar) * (yi - ybar) for xi, yi in zip(x, y))
-        sxx = fsum((xi - xbar) ** 2.0 for xi in x)
-        syy = fsum((yi - ybar) ** 2.0 for yi in y)
-        try:
-            return sxy / sqrt(sxx * syy)
-
-        except ZeroDivisionError:
-            raise statistics.StatisticsError('at least one of the inputs is constant') from None
+        return statistics.correlation(x, y)
 
     if issubclass(sectype, SecureFixedPoint):
         xbar = runtime.sum(x) / n
@@ -577,13 +556,6 @@ def correlation(x, y):
     raise TypeError('secure fixed-point type required')
 
 
-if sys.version_info.minor >= 10:
-    LinearRegression = statistics.LinearRegression
-else:
-    from collections import namedtuple
-    LinearRegression = namedtuple('LinearRegression', ('slope', 'intercept'))
-
-
 def linear_regression(x, y):
     """Return a (simple) linear regression model for x and y.
 
@@ -610,21 +582,7 @@ def linear_regression(x, y):
 
     sectype = type(x[0])  # all elts of x assumed of same type
     if not issubclass(sectype, SecureObject):
-        if sys.version_info.minor >= 10:
-            return statistics.linear_regression(x, y)
-
-        # inline code of statistics.linear_regression() adapted from Python 3.10.0:
-        xbar = fsum(x) / n
-        ybar = fsum(y) / n
-        sxy = fsum((xi - xbar) * (yi - ybar) for xi, yi in zip(x, y))
-        sxx = fsum((xi - xbar) ** 2.0 for xi in x)
-        try:
-            slope = sxy / sxx   # equivalent to:  covariance(x, y) / variance(x)
-        except ZeroDivisionError:
-            raise statistics.StatisticsError('x is constant') from None
-
-        intercept = ybar - slope * xbar
-        return LinearRegression(slope=slope, intercept=intercept)
+        return statistics.linear_regression(x, y)
 
     if issubclass(sectype, SecureFixedPoint):
         xbar = runtime.sum(x) / n
@@ -635,7 +593,7 @@ def linear_regression(x, y):
         sxx = runtime.in_prod(xxbar, xxbar)
         slope = sxy / sxx
         intercept = ybar - slope * xbar
-        return LinearRegression(slope=slope, intercept=intercept)
+        return statistics.LinearRegression(slope=slope, intercept=intercept)
 
     # TODO: implement for secure integers as well
     raise TypeError('secure fixed-point type required')

pyproject.toml

@@ -35,7 +35,7 @@ classifiers = [
     'Topic :: System :: Distributed Computing'
 ]
 license = {text = 'MIT License'}
-requires-python = '>=3.9'
+requires-python = '>=3.10'
 
 [project.urls]
 documentation = 'https://mpyc.readthedocs.io'

tests/test_mpctools.py

@@ -1,7 +1,7 @@
-import unittest
 import operator
 import itertools
 import functools
+import unittest
 from mpyc.runtime import mpc
 import mpyc.mpctools
 

tests/test_statistics.py

@@ -1,17 +1,11 @@
-import sys
-import unittest
 import random
 import statistics
+import unittest
 from mpyc.runtime import mpc
 from mpyc.statistics import (mean, variance, stdev, pvariance, pstdev,
                              mode, median, median_low, median_high, quantiles,
                              covariance, correlation, linear_regression)
 
-if sys.version_info.minor < 10:
-    statistics.covariance = covariance
-    statistics.correlation = correlation
-    statistics.linear_regression = linear_regression
-
 
 class Arithmetic(unittest.TestCase):