{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yx_wytHx0fDz"
      },
      "source": [
        "## $\\large{Quiz}$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YceKSrFj0fD4"
      },
      "source": [
        "請將 Path 每張影像讀取出來，並將影像整理成底下的 DataFrame 格式...<br>\n",
        "如下表 my_dataframe 及下圖 my_imageshow 的樣子"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "u6ftyJIt0fD4"
      },
      "outputs": [],
      "source": [
        "# upload Simpson.zip\n",
        "!wget -q https://github.com/TA-aiacademy/course_3.0/releases/download/CVCNN_Data/Simpson.zip\n",
        "# unzip file\n",
        "!unzip -q Simpson.zip"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "k_bkgCqc0fD5"
      },
      "outputs": [],
      "source": [
        "Path = \"./Simpson\""
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pvipLWYK0fD6"
      },
      "source": [
        "$\\begin{array}{c | c} class & label\\\\\\hline\n",
        " 0 & abraham\\_grampa\\_simpson\\\\\n",
        " 1 & agnes\\_skinner\\\\\n",
        " 2 & apu\\_nahasapeemapetilon\\\\\n",
        " 3 & bart\\_simpson\\\\\n",
        " 4 & carl\\_carlson \\end{array}$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OY4LUZf70fD6"
      },
      "source": [
        "## my_dataframe"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EO5KBohu0fD6"
      },
      "source": [
        "<img src=\"https://hackmd.io/_uploads/S17nWhPfT.png\" width=350/>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fiBDMyR50fD7"
      },
      "source": [
        "## my_imageshow"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wajRHowZ0fD7"
      },
      "source": [
        "![](https://hackmd.io/_uploads/SkHqW3PMa.png)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rqXXrxU20fD7"
      },
      "source": [
        "# ANS\n",
        "先觀察一下Simpson資料夾的結構<br>\n",
        "可以看出是\n",
        ">Simposon\n",
        ">>人物名稱\n",
        ">>>圖片名稱\n",
        "\n",
        "這次希望大家做的是用程式自動整理出一份含檔案位置、以及對應類別的表格<br>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JykA2rs70fD8"
      },
      "source": [
        "以這次小考的要求，主要可以分為四個步驟：\n",
        "1. [資料夾處理](#資料夾處理)\n",
        "2. [列舉檔案以及對應標籤](#列舉出所有檔案路徑以及對應的人物名稱)\n",
        "3. [DataFrame處理](#DataFrame處理)\n",
        "4. [畫圖呈現](#畫圖呈現)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7XmgYYiY0fD8"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import os\n",
        "import matplotlib.pyplot as plt"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fz6VPuNC0fD8"
      },
      "source": [
        "### 資料夾處理\n",
        "通常第一步要對資料夾處理時，會想要知道資料夾底下有什麼資料夾<br>\n",
        "而在 python 中使用 os.listdir(資料夾路徑) 可以得知指定的路徑底下有什麼檔案<br>\n",
        "還有一些常見作法可以參看以下討論區：<br>\n",
        "https://stackoverflow.com/questions/3207219/how-do-i-list-all-files-of-a-directory"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7Z1i5T2D0fD8"
      },
      "outputs": [],
      "source": [
        "folder = [os.path.join(Path, each) for each in os.listdir(Path)]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "E4MtXP980fD9"
      },
      "outputs": [],
      "source": [
        "folder"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZuSniRiS0fD9"
      },
      "source": [
        "有了每個人物對應的資料夾位置了後，現在該把每張圖片以及其對應的人物名稱整理出來"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "omCox68Z0fD9"
      },
      "source": [
        "### 列舉出所有檔案路徑以及對應的人物名稱"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yTW-oJgA0fD9"
      },
      "source": [
        "### Way1 -- os.listdir()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ge33gX6j0fD9"
      },
      "outputs": [],
      "source": [
        "datalist = []\n",
        "datalabel = []\n",
        "for root in folder:\n",
        "    for file in os.listdir(root):\n",
        "    # 由於使用 jupyter notebook時會產生一些附加檔案\n",
        "    # 但是我們不希望這些檔案被加進我們的資料之中，所以做了下列判斷式\n",
        "        if file.find('.ipynb_checkpoints') == -1:\n",
        "            datalist.append(os.path.join(root, file))\n",
        "            datalabel.append(root.split('/')[-1])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MSlmi_tu0fD-"
      },
      "source": [
        "### Way2 -- os.walk()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "32-g_YVZ0fD-"
      },
      "outputs": [],
      "source": [
        "datalist = []\n",
        "datalabel = []\n",
        "for roots, dirs, files in os.walk(Path):\n",
        "    if roots.find(\".ipynb_checkpoints\") == -1:\n",
        "        for file in files:\n",
        "            datalist.append(os.path.join(roots, file))\n",
        "            datalabel.append(root.split('/')[-1])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7437dhX00fD-"
      },
      "outputs": [],
      "source": [
        "print(*np.array((datalist, datalabel)).T, sep='\\n')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WCzpchlp0fD-"
      },
      "source": [
        "### DataFrame處理"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tt8M-Gmp0fD_"
      },
      "source": [
        "有了路徑以及名稱後，只要將其整理成DataFrame就大功告成了"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "sS6RSLyi0fD_"
      },
      "outputs": [],
      "source": [
        "data = {\"id_code\": datalist,\n",
        "        \"label\": datalabel}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1YApievO0fD_"
      },
      "outputs": [],
      "source": [
        "my_data = pd.DataFrame(data)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gunEUsl-0fD_"
      },
      "outputs": [],
      "source": [
        "classlabel = {'abraham_grampa_simpson': 0,\n",
        "              'agnes_skinner': 1,\n",
        "              'apu_nahasapeemapetilon': 2,\n",
        "              'bart_simpson': 3,\n",
        "              'carl_carlson': 4}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GB0h_GdQ0fD_"
      },
      "outputs": [],
      "source": [
        "my_data[\"label\"] = my_data[\"label\"].map(classlabel)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "HoSCYfXz0fD_"
      },
      "outputs": [],
      "source": [
        "my_data = my_data.sort_values('label')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "dgBBExSV0fD_"
      },
      "outputs": [],
      "source": [
        "my_data = my_data.reset_index(drop=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "t4HzO7vn0fEA"
      },
      "outputs": [],
      "source": [
        "my_data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "XOagjJTw0fEA"
      },
      "outputs": [],
      "source": [
        "my_data.to_csv(\"quiz1_ans_simspon.csv\", index=False)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "25SwBV5V0fEA"
      },
      "source": [
        "### 畫圖呈現"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "n4TaayFw0fEA"
      },
      "outputs": [],
      "source": [
        "plt.figure(figsize=(16, 20))\n",
        "for i in range(20):\n",
        "    ax = plt.subplot(5, 4, i+1)\n",
        "    plt.title(\"class {}\".format(my_data.loc[i, \"label\"]), fontsize=20)\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    ax.imshow(plt.imread(my_data.loc[i, \"id_code\"]))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZPp7z3-R0fEA"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "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": "3.7.12"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}