rf_scalability.ipynb

{
    "cells": [
        {
            "cell_type": "code",
            "execution_count": 1,
            "source": [
                "from joblib import cpu_count, parallel_backend\n",
                "\n",
                "\n",
                "print(f\"physical cores: {cpu_count(only_physical_cores=True)}\")"
            ],
            "outputs": [
                {
                    "output_type": "stream",
                    "name": "stdout",
                    "text": [
                        "physical cores: 8\n"
                    ]
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 14,
            "source": [
                "import platform\n",
                "\n",
                "\n",
                "platform.platform()"
            ],
            "outputs": [
                {
                    "output_type": "execute_result",
                    "data": {
                        "text/plain": [
                            "'macOS-11.5.1-arm64-i386-64bit'"
                        ]
                    },
                    "metadata": {},
                    "execution_count": 14
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 2,
            "source": [
                "from sklearn.datasets import make_classification\n",
                "from sklearn.ensemble import RandomForestClassifier\n",
                "\n",
                "\n",
                "X, y = make_classification(n_samples=int(1e4), n_features=50, random_state=0)\n",
                "clf = RandomForestClassifier(n_estimators=100, random_state=0)"
            ],
            "outputs": [],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 3,
            "source": [
                "%%timeit\n",
                "with parallel_backend(\"threading\", n_jobs=1):\n",
                "    clf.fit(X, y)"
            ],
            "outputs": [
                {
                    "output_type": "stream",
                    "name": "stdout",
                    "text": [
                        "4.38 s ± 24.2 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
                    ]
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 4,
            "source": [
                "%%timeit\n",
                "with parallel_backend(\"threading\", n_jobs=4):\n",
                "    clf.fit(X, y)"
            ],
            "outputs": [
                {
                    "output_type": "stream",
                    "name": "stdout",
                    "text": [
                        "1.17 s ± 20.8 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
                    ]
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 11,
            "source": [
                "4.38 / 1.17"
            ],
            "outputs": [
                {
                    "output_type": "execute_result",
                    "data": {
                        "text/plain": [
                            "3.7435897435897436"
                        ]
                    },
                    "metadata": {},
                    "execution_count": 11
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 5,
            "source": [
                "%%timeit\n",
                "with parallel_backend(\"threading\", n_jobs=8):\n",
                "    clf.fit(X, y)"
            ],
            "outputs": [
                {
                    "output_type": "stream",
                    "name": "stdout",
                    "text": [
                        "822 ms ± 8.61 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
                    ]
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 12,
            "source": [
                "4.38 / 0.822"
            ],
            "outputs": [
                {
                    "output_type": "execute_result",
                    "data": {
                        "text/plain": [
                            "5.328467153284672"
                        ]
                    },
                    "metadata": {},
                    "execution_count": 12
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 13,
            "source": [
                "%%timeit\n",
                "with parallel_backend(\"loky\", n_jobs=1):\n",
                "    clf.fit(X, y)"
            ],
            "outputs": [
                {
                    "output_type": "stream",
                    "name": "stdout",
                    "text": [
                        "4.34 s ± 10.2 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
                    ]
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 7,
            "source": [
                "%%timeit\n",
                "with parallel_backend(\"loky\", n_jobs=4):\n",
                "    clf.fit(X, y)"
            ],
            "outputs": [
                {
                    "output_type": "stream",
                    "name": "stdout",
                    "text": [
                        "1.36 s ± 22.3 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
                    ]
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": 8,
            "source": [
                "%%timeit\n",
                "with parallel_backend(\"loky\", n_jobs=8):\n",
                "    clf.fit(X, y)"
            ],
            "outputs": [
                {
                    "output_type": "stream",
                    "name": "stdout",
                    "text": [
                        "985 ms ± 32.2 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
                    ]
                }
            ],
            "metadata": {}
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "source": [],
            "outputs": [],
            "metadata": {}
        }
    ],
    "metadata": {
        "orig_nbformat": 4,
        "language_info": {
            "name": "python",
            "version": "3.9.2",
            "mimetype": "text/x-python",
            "codemirror_mode": {
                "name": "ipython",
                "version": 3
            },
            "pygments_lexer": "ipython3",
            "nbconvert_exporter": "python",
            "file_extension": ".py"
        },
        "kernelspec": {
            "name": "python3",
            "display_name": "Python 3.9.2 64-bit ('dev': conda)"
        },
        "interpreter": {
            "hash": "88e9b7219fe8fc278badea083ab1c022b86eb291352bb8a883adb96d8f084d4b"
        }
    },
    "nbformat": 4,
    "nbformat_minor": 2
}