S11_TriangleSchdulerCurve.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Welcome to Colaboratory",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/thamizhannal/eva5/blob/master/S11_TriangleSchdulerCurve.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u3joIIBoooBK"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "#from curlyBrace import curlyBrace\n",
        "import os\n",
        "\n",
        "def clr(lr_min,lr_max,x):\n",
        "  \"\"\" return lr rate at a x \"\"\"\n",
        "  lr = lr_min + (lr_max-lr_min) * (1-x)\n",
        "  return lr\n",
        "\n",
        "def fn_x(iteration, stepsize):\n",
        "  \"\"\" get x for iteration at stepsize\"\"\"\n",
        "  cycle = np.floor(1+(iteration/stepsize)/2)\n",
        "  temp = (iteration/stepsize) - 2* cycle + 1\n",
        "  x = np.abs(temp)\n",
        "  return x"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8jRd0ysZCdiq",
        "outputId": "6860f151-b423-40a4-a340-0e7b692f2160",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        }
      },
      "source": [
        "! git clone https://github.com/iruletheworld/matplotlib-curly-brace"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Cloning into 'matplotlib-curly-brace'...\n",
            "remote: Enumerating objects: 357, done.\u001b[K\n",
            "remote: Total 357 (delta 0), reused 0 (delta 0), pack-reused 357\u001b[K\n",
            "Receiving objects: 100% (357/357), 8.33 MiB | 29.92 MiB/s, done.\n",
            "Resolving deltas: 100% (149/149), done.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DEBkrpCq9W36"
      },
      "source": [
        "def bracket(ax, pos=[0,0], scalex=1, scaley=1, text=\"\",textkw = {}, linekw = {}):\n",
        "    x = np.array([0, 5, 990,1000])\n",
        "    y = np.array([0,-0.05,-0.05,-0.1])\n",
        "    x = np.concatenate((x,x+0.5)) \n",
        "    y = np.concatenate((y,y[::-1]))\n",
        "    ax.plot(x*scalex+pos[0], y*scaley+pos[1], clip_on=False, \n",
        "            transform=ax.get_xaxis_transform(), **linekw)\n",
        "    ax.text(pos[0]+0.5*scalex, (y.min()-0.01)*scaley+pos[1], text, \n",
        "                transform=ax.get_xaxis_transform(),\n",
        "                ha=\"center\", va=\"top\", **textkw)"
      ],
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dxqYApyFoof0",
        "outputId": "a545b928-736a-45fc-a2e3-c91ef24d76e4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        }
      },
      "source": [
        "\n",
        "x = []\n",
        "y = [] \n",
        "\n",
        "for i in range(1,2000):\n",
        "  t = fn_x(iteration=i, stepsize=500)\n",
        "  x.append(t)\n",
        "  y.append(clr(lr_min=.01,lr_max=.1,x=t))\n",
        "\n",
        "plt.title(\"Triangular Schedule\")\n",
        "plt.text(0, 0.01, r'min_lr')\n",
        "plt.text(0, 0.10, r'max_lr')\n",
        "plt.text(0, 0.01, r'min_lr')\n",
        "plt.plot(range(1,2000),y)\n"
      ],
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Triangular Schedule')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yigdzTzQosNn"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}