rename pacakge to pyrandvec; add environment.yml for (Ana)conda

MANIFEST.in

@@ -2,4 +2,5 @@ exclude visualise.py
 exclude setup.cfg
 exclude Makefile
 exclude publish.sh
+exclude environment.yml
 exclude pyproject.toml_bck

Makefile

@@ -47,3 +47,6 @@ test:
 flake8:
 	flake8
 
+conda:
+	conda env create -f environment.yml
+	conda activate pyrandvec

environment.yml

@@ -0,0 +1,21 @@
+name: pyrandvec
+channels:
+  - conda-forge
+  - defaults
+dependencies:
+  # Basic conda install
+  - python=3.9
+  - matplotlib
+  - codecov
+  - pip
+  - pip:
+    - --editable . # install locally
+    # Testing
+    - pytest
+    # Linting
+    - flake8==5.0.4
+    - flake8-quotes==3.3.1
+    - flake8-bandit==4.1.1
+    - pep8-naming==0.13.2
+    - flake8-docstrings==1.6.0
+    - flake8-use-fstring==1.4.0

randvec.py → pyrandvec.py

@@ -1,11 +1,11 @@
+"""Module with methods for the generation of random vectors with a fixed sum of one."""
 import random
 import math
 
 
-def sample(n: int, d: int, method: str = "normalisation", shuffle: bool = False) -> list[list[float]]:
-    '''
-    Generates a list of n d-dimensional random vectors whose elements sum up to one
-    with different methods.
+def sample(n: int, d: int, method: str = 'normalisation', shuffle: bool = False) -> list[list[float]]:
+    """
+    Generate a list of n d-dimensional random vectors whose elements sum up to one.
 
     Args:
         n (int): desired number of vectors.
@@ -14,7 +14,7 @@ def sample(n: int, d: int, method: str = "normalisation", shuffle: bool = False)
         shuffle (bool): shall each vector be randomly shuffled? Default is False.
     Returns:
         List of length n of d-dimensional lists.
-    '''
+    """
     assert n >= 1
     assert d >= 2
 
@@ -37,30 +37,33 @@ def sample(n: int, d: int, method: str = "normalisation", shuffle: bool = False)
 
 
 def normalise(x: list[float]) -> list[float]:
-    '''
-    Utility function to normalise a list of floating point numbers.
-    I.e., the function divides each component of the list by its sum.
+    """
+    Normalise a list of floating point numbers.
+
+    I.e., the function divides each component of the list by its sum such that
+    the resulting list has sum one.
 
     Args:
         x (list[float]): Input list.
     Returns:
         Normalised list.
-    '''
+    """
     return list(map(lambda e: e / sum(x), x))
 
 
 def sample_normalisation(n: int, d: int) -> list[list[float]]:
-    '''
-    Generates a list of random vectors via normalisation.
-    I.e., (1) each component is sampled from a U(0,1) distribution and subsequently
-    (2) each component is divided by the components' sum.
+    """
+    Generate a list of random vectors via normalisation.
+
+    The process works as follows: (1) each component is sampled from a U(0,1) distribution
+    and subsequently (2) each component is divided by the components' sum.
 
     Args:
         n (int): desired number of vectors.
         d (int): dimension.
     Returns:
         List of length n of d-dimensional lists.
-    '''
+    """
     assert n >= 1
     assert d >= 2
 
@@ -74,9 +77,10 @@ def sample_normalisation(n: int, d: int) -> list[list[float]]:
 
 
 def sample_iterative(n: int, d: int) -> list[list[float]]:
-    '''
-    Generates a list of random vectors via an iterative approach:
-    I.e., the i-th component of the rpv is sampled uniformly at random from [0, s] where s is
+    """
+    Generate a list of random vectors via an iterative approach.
+
+    More precisely, the i-th component of the rpv is sampled uniformly at random from [0, s] where s is
     the sum of the 0, ..., (i-1)st components. The last component is finally (1-s). This
     way it is unsured that the vectors are normalised.
 
@@ -85,7 +89,7 @@ def sample_iterative(n: int, d: int) -> list[list[float]]:
         d (int): dimension.
     Returns:
         List of length n of d-dimensional lists.
-    '''
+    """
     assert n >= 1
     assert d >= 2
 
@@ -103,8 +107,9 @@ def sample_sample_iterative(d):
 
 
 def sample_trigonometric(n: int, d: int) -> list[list[float]]:
-    '''
-    Generates a list of random vectors via a trigonometric method.
+    """
+    Generate a list of random vectors via a trigonometric method.
+
     Section 5 in the following paper contains the details: Maziero, J. Generating Pseudo-Random
     Discrete Probability Distributions. Brazilian Journal of Physics 45, 377–382 (2015).
     https://doi.org/10.1007/s13538-015-0337-8)
@@ -114,7 +119,7 @@ def sample_trigonometric(n: int, d: int) -> list[list[float]]:
         d (int): dimension.
     Returns:
         List of length n of d-dimensional lists.
-    '''
+    """
     assert n >= 1
     assert d >= 2
 
@@ -141,9 +146,10 @@ def sample_sample_trigonometric(d):
 
 
 def sample_exponential(n: int, d: int) -> list[list[float]]:
-    '''
-    Generates a list of random vectors by means of the inverse
-    exponential distribution function: I.e., the i-th component of the rpv is sampled
+    """
+    Generate a list of random vectors by means of the inverse exponential distribution function.
+
+    More precisely, the i-th component of the rpv is sampled
     uniformly at random from [0, s] where s is the sum of the 0, ..., (i-1)st
     components. The last component is finally (1-s). This way it is unsured that
     the vectors are normalised.
@@ -153,7 +159,7 @@ def sample_exponential(n: int, d: int) -> list[list[float]]:
         d (int): dimension.
     Returns:
         List of length n of d-dimensional lists.
-    '''
+    """
     assert n >= 1
     assert d >= 2
 
@@ -166,8 +172,8 @@ def sample_exponential(n: int, d: int) -> list[list[float]]:
 
 
 def sample_simplex(n: int, d: int) -> list[list[float]]:
-    '''
-    Generates a list of random vectors via simplex sampling.
+    """
+    Generate a list of random vectors via simplex sampling.
 
     See the following paper for details on this method:
     Grimme, C. Picking a Uniformly Random Point from an Arbitrary Simplex.
@@ -178,7 +184,7 @@ def sample_simplex(n: int, d: int) -> list[list[float]]:
         d (int): dimension.
     Returns:
         List of length n of d-dimensional lists.
-    '''
+    """
     assert n >= 1
     assert d >= 2
 

setup.cfg

@@ -6,6 +6,7 @@ exclude =
   .pytest_cache,
   .mypy_cache,
   docs,
+  tests
 ignore =
   # unexpected spaces around keyword / parameter equal
   # Allows for param = value instead of enforcing param=value

setup.py

@@ -1,46 +1,46 @@
 import setuptools
 
-with open("README.md", "r") as fh:
+with open('README.md', 'r') as fh:
     long_description = fh.read()
 
 setuptools.setup(
-    name = "randvec",
-    version = "1.0.2",
-    author = "Jakob Bossek",
-    author_email = "j.bossek@gmail.com",
-    description = "Generate random and uniform vectors whose components sum up to one",
+    name = 'pyrandvec',
+    version = '1.0.2',
+    author = 'Jakob Bossek',
+    author_email = 'j.bossek@gmail.com',
+    description = 'Generate random vectors whose components sum up to one',
     long_description = long_description,
-    url = "https://github.com/jakobbossek/randvec/",
+    url = 'https://github.com/jakobbossek/pyrandvec/',
     # tell PyPi that we use markdown and not RestructuredText
-    long_description_content_type = "text/markdown",
-    packages = setuptools.find_packages(exclude = ["tests"]),
+    long_description_content_type = 'text/markdown',
+    packages = setuptools.find_packages(exclude = ['tests']),
     python_requires = '>=3.7',
-    py_modules = ["randvec"],
+    py_modules = ['pyrandvec'],
     package_dir = {},
     # make it easy for users to find the package
     # (see https://pypi.org/classifiers/)
     classifiers = [
-        "Operating System :: OS Independent",
-        "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.6",
-        "Programming Language :: Python :: 3.7",
-        "Programming Language :: Python :: 3.8",
-        "Programming Language :: Python :: 3.9",
-        "Programming Language :: Python :: 3.10",
-        "Programming Language :: Python :: 3.11",
-        "License :: OSI Approved :: MIT License",
-        "Development Status :: 5 - Production/Stable",
-        "Intended Audience :: Developers",
-        "Intended Audience :: Science/Research",
-        "Topic :: Scientific/Engineering",
-        "Topic :: Scientific/Engineering :: Mathematics",
+        'Operating System :: OS Independent',
+        'Programming Language :: Python :: 3',
+        'Programming Language :: Python :: 3.6',
+        'Programming Language :: Python :: 3.7',
+        'Programming Language :: Python :: 3.8',
+        'Programming Language :: Python :: 3.9',
+        'Programming Language :: Python :: 3.10',
+        'Programming Language :: Python :: 3.11',
+        'License :: OSI Approved :: MIT License',
+        'Development Status :: 5 - Production/Stable',
+        'Intended Audience :: Developers',
+        'Intended Audience :: Science/Research',
+        'Topic :: Scientific/Engineering',
+        'Topic :: Scientific/Engineering :: Mathematics',
     ],
     # for production dependencies
     install_requires = [],
     # for optional (development) requirements
     extras_require = {
-        "dev": [
-            "pytest >=3.7",
+        'dev': [
+            'pytest >=3.7',
         ]
     },
 )

tests/test_sample.py

@@ -1,11 +1,11 @@
 import itertools
-from randvec import sample
+from pyrandvec import sample
 
 
 def test_that_vector_sum_is_one():
     n = 5
     ds = [2, 4]
-    methods = ["normalisation", "iterative", "exponential", "simplex", "trigonometric"]
+    methods = ['normalisation', 'iterative', 'exponential', 'simplex', 'trigonometric']
     shuffling = [True, False]
 
     for d, method, shuffle in list(itertools.product(ds, methods, shuffling)):

visualise.py

@@ -1,13 +1,13 @@
 import matplotlib.pyplot as plt
-from randvec import sample
+from pyrandvec import sample
 
 # parameters
 n, d = 1000, 3
 
 # plot
 fig = plt.figure()
 
-methods = ["normalisation", "iterative", "exponential", "simplex", "trigonometric"]
+methods = ['normalisation', 'iterative', 'exponential', 'simplex', 'trigonometric']
 k = 1
 for i, method in enumerate(methods):
     for j, shuffle in enumerate([True, False]):
@@ -28,7 +28,7 @@
         sp.set_zlabel('z')
         title = method
         if shuffle:
-            title += " (shuffled)"
+            title += ' (shuffled)'
         sp.set_title(title)
         sp.scatter(xs, ys, zs, marker = 'o')