v0.2.38 updates:

* Lectures notebooks CLI test.
* Version string without exec in setup.py.
* Development sections in README.

Author: msuzen

.gitignore

@@ -119,3 +119,4 @@ Icon[]
 Network Trash Folder
 Temporary Items
 .apdisk
+lecturePy

NEWS

@@ -1,3 +1,9 @@
+v0.2.38
+
+* `runlectures.sh` is added for running lectures notebooks on CLI for testing. 
+* Update in README, new section about developments.
+* Update how version is updated in `setup.py` dropping `exec` like practice.
+
 v0.2.33
 
 * `mean_adjacent_gap_ratio`, return ratios matrix as well.

README.md

@@ -26,11 +26,9 @@ specific release.
 
 ## Approach and features
 
-The package provides tools and utilities for randomness based research with `High-Entropy Random Number Generation (HE-RNG)`. It means generation iss performed with non-deterministic seeds every time a random
+The package provides tools and utilities for randomness based research with `High-Entropy Random Number Generation (HE-RNG)`. It means generation is performed with non-deterministic seeds every time a random
 library function is called. 
 
-There is a common misconception in computational sciences that speed is the ultimate goal, however primary objective is scientific correctness first. For this reasons, scientific correctness is taken precedence over speed in the development of the package. For proven methods being a baseline, we might implement faster versions. 
-
 ### High-entropy random number utilities 
 
 The core package is providing strong randomness improving the simulation quality. We use NumPy grammar and as a backend.
@@ -69,6 +67,16 @@ Lectures notes that introduce randomization concepts with the usage of `Leymosun
 * [he_rng_nist.ipynb](https://github.com/msuzen/leymosun/blob/main/lectures/he_rng_nist.ipynb): `Lecture on Understanding High-Entropy RNGs with NIST  benchmark`. This lecture provides a way to test different RNGs or usage of RNGs via standard quality 
 tests. 
 
+## Development notes 
+* Philosophy  
+  There is a common misconception in computational sciences that speed is the ultimate goal, however primary objective is scientific correctness first. For this reasons, scientific correctness is taken precedence over speed in the development of the package. For proven methods being a baseline, we might implement faster versions. 
+* Testing 
+  * `tests` and `nbconvert` should be present as recommended dependency.
+  * Test script should pass before any release. 
+    Unit tests `runtests.sh` and lectures `runlectures.sh`. (`lecturePy` directory is needed but this is ignored in the repo via `.gitignore`).
+  * Add unit tests for each new method and features. 
+  * Add run portion for the new lecture in `runlecture.sh`.
+
 ## Publications
 
 Papers, datasets and other material that used `leymosun`. 

lectures/wigner_cats.ipynb

@@ -42,7 +42,7 @@
     "import numpy as np\n",
     "import matplotlib.pyplot as plt\n",
     "from leymosun.gaussian import goe, wigner\n",
-    "from leymosun.matrix import mixed_ensemble \n",
+    "from leymosun.matrix import mixed_ensemble\n",
     "from leymosun.spectral import empirical_spectral_density\n",
     "from leymosun.stats import bootstrap_observed_matrix_ci\n",
     "import leymosun\n",

lectures/wigner_semicircle.ipynb

@@ -190,7 +190,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "dwigner = wigner(locations, domasigmin_boundary=2.0)"
+    "dwigner = wigner(locations, domain_boundary=2.0)"
    ]
   },
   {

leymosun/gaussian.py

@@ -25,7 +25,7 @@ def goe(matrix_order: int, scale: float = 1.0, loc: float= 0.0, bitget: BitGener
     return 0.50 * (A + A.transpose())
 
 
-def wigner(locations: np.array, domasigmin_boundary=2.0):
+def wigner(locations: np.array, domain_boundary=2.0):
     """Wigner semi-circle density for GOE
 
     Density is defined as
@@ -39,7 +39,7 @@ def wigner(locations: np.array, domasigmin_boundary=2.0):
         The density at the locations, list.
 
     """
-    domain_boundary_sqr = domasigmin_boundary**2
+    domain_boundary_sqr = domain_boundary**2
     return [
         2.0
         * np.sqrt(np.abs(domain_boundary_sqr - x**2))

leymosun/version.py

@@ -1 +1 @@
-__version__="0.2.33"
+__version__="0.2.38"

runlectures.sh

@@ -0,0 +1,45 @@
+#!/bin/bash
+#
+# Run notebooks after nbconvert
+#
+# Set -e: exit immediately if any command fails
+set -e
+
+# Set -u: error if undefined variable used
+set -u
+
+# Set -o pipefail: fail on pipeline failure (e.g., command fails in pipe)
+set -o pipefail
+
+# wigner_cats.ipynb
+echo " "
+echo " Running wigner_cats.ipynb"
+echo " "
+rm -f lecturePy/wigner_cats.py
+jupyter nbconvert --to python lectures/wigner_cats.ipynb --output-dir=./lecturePy --log-level=ERROR
+python lecturePy/wigner_cats.py
+echo " "
+echo " Success running wigner_cats.ipynb"
+echo " "
+
+# wigner_dyson_spacing.ipynb
+echo " "
+echo " Running wigner_dyson_spacing.ipynb"
+echo " "
+rm -f lecturePy/wigner_dyson_spacing.py
+jupyter nbconvert --to python lectures/wigner_dyson_spacing.ipynb --output-dir=./lecturePy --log-level=ERROR
+python lecturePy/wigner_dyson_spacing.py
+echo " "
+echo " Success running wigner_dyson_spacing.ipynb"
+echo " "
+
+# wigner_semicircle.ipynb
+echo " "
+echo " Running wigner_semicircle.ipynb"
+echo " "
+rm -f lecturePy/wigner_semicircle.py
+jupyter nbconvert --to python lectures/wigner_semicircle.ipynb --output-dir=./lecturePy --log-level=ERROR
+python lecturePy/wigner_semicircle.py
+echo " "
+echo " Success running wigner_semicircle.ipynb"
+echo " "

setup.py

@@ -1,7 +1,8 @@
 from setuptools import setup
 
 fp = open("./leymosun/version.py")
-exec(fp.read())
+sversion = fp.read()
+__version__ = sversion.split('=')[1][1:-1]
 
 with open("README.md") as f:
     long_description = f.read()
@@ -34,7 +35,7 @@
     ],
     test_suite="tests",
     extras_require={
-        "test":["pytest==8.4.2", "coverage==7.10.7", "pytest-cov==7.0.0"]
+        "test":["pytest==8.4.2", "coverage==7.10.7", "pytest-cov==7.0.0", "nbconvert==7.16.6"]
     },
     zip_safe=False,
 )