v2.2.0dev0

.github/workflows/pytest-on-each-version.yaml

@@ -2,15 +2,18 @@
 # - https://enu23456.hatenablog.com/entry/2022/11/24/195744
 name: Test on each version
 
-on: [push]
+on:
+  push:
+    branches:
+      - 'main'
 
 jobs:
   build-linux:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: ["3.7", "3.8", "3.9", "3.10", "3.11"]
-      max-parallel: 5
+        python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12"]
+      max-parallel: 6
 
     steps:
     - uses: actions/checkout@v3

README.md

@@ -15,7 +15,7 @@
 This is an algorithm for evenly partitioning data in a `scikit-learn`-like interface.
 (See [References](#References) for details of the algorithm.)
 
-![simulation_gif](https://github.com/yu9824/kennard_stone/blob/main/example/simulate.gif?raw=true "simulation_gif")
+![simulation_gif](https://github.com/yu9824/kennard_stone/blob/main/examples/simulate.gif?raw=true "simulation_gif")
 
 ## How to install
 
@@ -41,7 +41,7 @@ You need `numpy>=1.20` and `scikit-learn` to run.
 
 You can use them like [scikit-learn](https://scikit-learn.org/stable/modules/classes.html#module-sklearn.model_selection).
 
-See [example](https://github.com/yu9824/kennard_stone/tree/main/example) for details.
+See [examples](https://github.com/yu9824/kennard_stone/tree/main/examples) for details.
 
 In the following, `X` denotes an arbitrary explanatory variable and `y` an arbitrary objective variable.
 And, `estimator` indicates an arbitrary prediction model that conforms to scikit-learn.
@@ -135,7 +135,7 @@ print(cross_validate(estimator, X, y, cv=5))
 There is no notion of `random_state` or `shuffle` because the partitioning is determined uniquely for the dataset.
 If these arguments are included, they do not cause an error. They simply have no effect on the result. Please be careful.
 
-If you want to run the notebook in example directory,
+If you want to run the notebook in examples directory,
 you will need to additionally install `pandas`, `matplotlib`, `seaborn`, `tqdm`, and `jupyter` other than the packages in requirements.txt.
 
 ## Distance metrics
@@ -252,3 +252,8 @@ Copyright (c) 2021 yu9824
 
 - Improve typing in `kennard_stone.train_test_split`
 - Add some docstrings.
+
+### v2.2.0
+
+- Supports GPU calculations. (when metric is 'euclidean', 'manhattan', 'chebyshev' and 'minkowski')
+- Supports Python 3.12

examples/cpu_vs_gpu.ipynb

@@ -0,0 +1,151 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# CPU vs GPU"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Device info\n",
+    "\n",
+    "- https://support.apple.com/kb/SP825"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from time import perf_counter\n",
+    "\n",
+    "from sklearn.datasets import fetch_california_housing\n",
+    "\n",
+    "# You need to install 'pytorch' when you use GPU\n",
+    "import torch\n",
+    "\n",
+    "from kennard_stone import train_test_split\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(20640, 8)"
+      ]
+     },
+     "execution_count": 2,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "X = fetch_california_housing().data\n",
+    "X.shape\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## CPU"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Calculating pairwise distances using scikit-learn.\n",
+      "Calculating pairwise distances using scikit-learn.\n",
+      "CPU: 178.731 sec\n"
+     ]
+    }
+   ],
+   "source": [
+    "t_start = perf_counter()\n",
+    "_ = train_test_split(\n",
+    "    X, test_size=0.2, n_jobs=-1, device=\"cpu\", metric=\"euclidean\"\n",
+    ")\n",
+    "t_end = perf_counter()\n",
+    "t_cpu = t_end - t_start\n",
+    "print(f\"CPU: {t_cpu:.3f} sec\")\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## GPU"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Calculating pairwise distances using PyTorch on 'mps'.\n",
+      "Calculating pairwise distances using PyTorch on 'mps'.\n",
+      "mps: 33.525 sec\n"
+     ]
+    }
+   ],
+   "source": [
+    "# cuda\n",
+    "if torch.cuda.is_available():\n",
+    "    device = \"cuda\"\n",
+    "# mps (M1 Mac)\n",
+    "elif torch.backends.mps.is_available():\n",
+    "    device = \"mps\"\n",
+    "else:\n",
+    "    raise RuntimeError(\"GPU is not available.\")\n",
+    "\n",
+    "t_start = perf_counter()\n",
+    "_ = train_test_split(\n",
+    "    X, test_size=0.2, n_jobs=-1, device=device, metric=\"euclidean\"\n",
+    ")\n",
+    "t_end = perf_counter()\n",
+    "t_gpu = t_end - t_start\n",
+    "print(f\"{device}: {t_gpu:.3f} sec\")\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "kennard-stone",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

kennard_stone/__init__.py

@@ -7,7 +7,7 @@
 
 from .kennard_stone import KFold, train_test_split
 
-__version__ = "2.1.6"
+__version__ = "2.2.0dev0"
 __license__ = "MIT"
 __author__ = "yu9824"
 __copyright__ = "Copyright © 2021 yu9824"