{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Анализируем лица с помощью искусственного интеллекта\n",
    "\n",
    "Материалы воркшопа с зимней школы **Искусственный интеллект в психологии 2020**.\n",
    "\n",
    "В современной науке всё большее значение принимают методы, основанные на машинном обучении и анализе данных. Мы попробуем использовать анализ данных для того, чтобы сделать некоторые (исключительно грубые) выводы о человеческой психологии на основа анализа изображений.\n",
    "\n",
    "В нашем случае мы будем использовать небольшое количество изображений, для ускорения процесса обработки. Поэтому выводы, которые мы сможем сделать, не будут статистически обоснованы. Однако в случае, если в нашем распоряжении окажется большее количество фотографий, мы сможем попробовать получить более достоверные результаты с помощью такой же методики."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Когнитивные сервисы для анализа изображений\n",
    "\n",
    "Большое количество контента в интернет/соц.сетях - это изображения. Мы можем использовать нейросетевые модели для извлечения из изображений глубинной информации о содержимом, например:\n",
    " * Для каждого из людей на фото - пол, возраст, испытываемые эмоции, наличие аксессуаров на лице\n",
    " * Список объектов на фотографии\n",
    " * Контекст: в помещении, на улице\n",
    " * Тип фото: портрет, групповое фото, пейзаж\n",
    " * Цветовая гамма\n",
    " \n",
    "В нашем случае мы сконцентрируемся на извлечении информации о людях. Для этого будем использовать облачный сервис [Microsoft Azure Face API](https://docs.microsoft.com/azure/cognitive-services/face/overview/?wt.mc_id=psycho-events-dmitryso).\n",
    "\n",
    "Для использования Face API нам необходим специальный ключ, а также интернет-адрес для вызова сервиса (Endpoint URL). Есть несколько способов получить ключ:\n",
    "\n",
    "* Если у Вас есть подписка Microsoft Azure, [создайте ресурс Face API](https://docs.microsoft.com/azure/cognitive-services/cognitive-services-apis-create-account/?wt.mc_id=crpyface-github-dmitryso), и скопируйте ключ и URL оттуда.\n",
    "* Вы всегда можете [создать бесплатную пробную подписку](https://azure.microsoft.com/free/?wt.mc_id=psycho-events-dmitryso), по которой в течение месяца вы сможете расходовать определенный предоставленный вам кредитный лимит, а затем в течение года получать ограниченный набор сервисов бесплатно (для этого понадобится кредитная карта).\n",
    "* Студенты могут завести облачный аккаунт [здесь](http://aka.ms/az4stud), верифицировав свой студенческий статус с помощью e-mail в домене университета. Альтернативно, сразу много бесплатных студенческих ресурсов можно получить через [GitHub Student Developer Pack](https://education.github.com/pack) - он поддерживает верификацию по документу, но она требует чуть больше времени. \n",
    "\n",
    "**Важно:** При использовании бесплатной категории Face API, будут некоторые ограничения по количеству изображений, которые вы сможете обработать (несколько тысяч), а также по частоте запросов (не более 20 запросов в минуту). Для обеспечения работы без ошибок, возможно, придется вставлять паузы между запросами к сервису.\n",
    "\n",
    "Полученный ключ и адрес Endpoint URL введите в ячейке ниже:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "key = '--INSERT YOUR KEY HERE--'\n",
    "endpoint = 'https://westus2.api.cognitive.microsoft.com' \n",
    "# You might need different endpoint, check the cognitive services page!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Для работы с Face API нам нужно установить специальную библиотеку, а также убедиться, что другие важные библиотеки Python установлены:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "!{sys.executable} -m pip install --quiet --user matplotlib pandas ipyplot\n",
    "!{sys.executable} -m pip install --quiet --user azure-cognitiveservices-vision-face\n",
    "import os, matplotlib.pyplot as plt, json, pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Все вызовы к Face API производятся с помощью модуля `cognitive_face`, который мы для краткости назовём `cf`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import azure.cognitiveservices.vision.face as cf\n",
    "from msrest.authentication import CognitiveServicesCredentials\n",
    "cli = cf.FaceClient(endpoint,CognitiveServicesCredentials(key))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Основные функции для обработки лица -- `face.detect_with_url` и `face.detect_with_stream`. Они извлекают из лица разную информацию в зависимости от передаваемых параметров. Например. попробуем проанализировать следующее изображение:\n",
    "![](https://2016.dotnext-piter.ru/assets/images/people/soshnikov.jpg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'face_rectangle': {'width': 127, 'height': 127, 'left': 83, 'top': 87}, 'face_attributes': {'age': 39.0, 'gender': 'male', 'emotion': {'anger': 0.0, 'contempt': 0.0, 'disgust': 0.0, 'fear': 0.0, 'happiness': 1.0, 'neutral': 0.0, 'sadness': 0.0, 'surprise': 0.0}}}\n"
     ]
    }
   ],
   "source": [
    "res = cli.face.detect_with_url('https://2016.dotnext-piter.ru/assets/images/people/soshnikov.jpg',\n",
    "                               return_face_attributes=['emotion','age','gender'], return_face_id=False)\n",
    "print(res[0].as_dict())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Анализируем фотографии с летней школы\n",
    "\n",
    "Проанализируем фотографии с летней школы ИИ, расположенные в интернет [вот тут](http://psy.msu.ru/science/conference/summer/2019/photo/index.html).\n",
    "Можно заметить, что все фотографии имеют адрес http://psy.msu.ru/science/conference/summer/2019/photo/xxx.jpg, где xxx - номер фото, от 000 до 131."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "urls = [\"http://psy.msu.ru/science/conference/summer/2019/photo/{:03d}.jpg\".format(i) for i in range(1,131)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import ipyplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">0</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/001.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/001.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">1</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/011.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/011.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">2</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/021.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/021.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">3</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/031.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/031.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">4</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/041.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/041.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">5</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/051.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/051.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">6</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/061.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/061.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">7</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/071.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/071.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">8</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/081.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/081.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">9</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/091.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/091.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">10</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/101.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/101.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">11</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/111.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/111.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div><div style=\"display: inline-block; width: 170px; vertical-align: top; text-align: center;\"><h4 style=\"font-size: 12px\">12</h4><h4 style=\"font-size: 9px; padding-left: 10px; padding-right: 10px; width: 90%; word-wrap: break-word; white-space: normal;\">http://psy.msu.ru/science/conference/summer/2019/photo/121.jpg</h4><img src=\"http://psy.msu.ru/science/conference/summer/2019/photo/121.jpg\" style=\"margin: 1px; width: 150px; border: 2px solid #ddd;\"/></div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ipyplot.plot_images(urls[::10],img_width=150)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/001.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/002.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/003.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/004.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/005.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/006.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/007.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/008.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/009.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/010.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/011.jpg ...4  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/012.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/013.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/014.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/015.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/016.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/017.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/018.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/019.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/020.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/021.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/022.jpg ...4  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/023.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/024.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/025.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/026.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/027.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/028.jpg ...8  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/029.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/030.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/031.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/032.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/033.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/034.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/035.jpg ...4  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/036.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/037.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/038.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/039.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/040.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/041.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/042.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/043.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/044.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/045.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/046.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/047.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/048.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/049.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/050.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/051.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/052.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/053.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/054.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/055.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/056.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/057.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/058.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/059.jpg ...4  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/060.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/061.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/062.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/063.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/064.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/065.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/066.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/067.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/068.jpg ...4  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/069.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/070.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/071.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/072.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/073.jpg ...4  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/074.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/075.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/076.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/077.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/078.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/079.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/080.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/081.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/082.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/083.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/084.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/085.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/086.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/087.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/088.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/089.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/090.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/091.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/092.jpg ...3  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/093.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/094.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/095.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/096.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/097.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/098.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/099.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/100.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/101.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/102.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/103.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/104.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/105.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/106.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/107.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/108.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/109.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/110.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/111.jpg ...4  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/112.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/113.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/114.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/115.jpg ...12  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/116.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/117.jpg ...0  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/118.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/119.jpg ...1  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/120.jpg ...10  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/121.jpg ...12  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/122.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/123.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/124.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/125.jpg ...2  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/126.jpg ...5  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/127.jpg ...6  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/128.jpg ...8  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/129.jpg ...7  faces\n",
      "Analyzing http://psy.msu.ru/science/conference/summer/2019/photo/130.jpg ...9  faces\n",
      "Found 333 faces\n"
     ]
    }
   ],
   "source": [
    "data = []\n",
    "for x in urls:\n",
    "    print(\"Analyzing {} ...\".format(x),end='')\n",
    "    res = cli.face.detect_with_url(x, return_face_attributes=['emotion','age','gender'], return_face_id=False)\n",
    "    nfaces = len(res)\n",
    "    print(nfaces, ' faces')\n",
    "    for t in res:\n",
    "        fa = t.as_dict()['face_attributes']\n",
    "        em = fa['emotion']\n",
    "        z = { 'gender' : fa['gender'], 'age' : fa['age'], 'nfaces' : nfaces }\n",
    "        data.append(dict(list(z.items())+list(em.items())))\n",
    "print(\"Found {} faces\".format(len(data)))\n",
    "\n",
    "with open('data.json','w') as f:\n",
    "    json.dump(data,f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>nfaces</th>\n",
       "      <th>anger</th>\n",
       "      <th>contempt</th>\n",
       "      <th>disgust</th>\n",
       "      <th>fear</th>\n",
       "      <th>happiness</th>\n",
       "      <th>neutral</th>\n",
       "      <th>sadness</th>\n",
       "      <th>surprise</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>female</td>\n",
       "      <td>24.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.991</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>23.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "      <td>23.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>24.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>female</td>\n",
       "      <td>25.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>328</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>329</td>\n",
       "      <td>female</td>\n",
       "      <td>28.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.904</td>\n",
       "      <td>0.096</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>330</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.647</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>331</td>\n",
       "      <td>female</td>\n",
       "      <td>18.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.864</td>\n",
       "      <td>0.136</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>332</td>\n",
       "      <td>female</td>\n",
       "      <td>28.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>333 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     gender   age  nfaces  anger  contempt  disgust  fear  happiness  neutral  \\\n",
       "0    female  24.0       1    0.0     0.000      0.0   0.0      0.991    0.009   \n",
       "1    female  23.0       1    0.0     0.000      0.0   0.0      1.000    0.000   \n",
       "2    female  23.0       3    0.0     0.000      0.0   0.0      1.000    0.000   \n",
       "3    female  24.0       3    0.0     0.000      0.0   0.0      1.000    0.000   \n",
       "4    female  25.0       3    0.0     0.000      0.0   0.0      1.000    0.000   \n",
       "..      ...   ...     ...    ...       ...      ...   ...        ...      ...   \n",
       "328  female  19.0       9    0.0     0.000      0.0   0.0      1.000    0.000   \n",
       "329  female  28.0       9    0.0     0.000      0.0   0.0      0.904    0.096   \n",
       "330  female  22.0       9    0.0     0.001      0.0   0.0      0.350    0.647   \n",
       "331  female  18.0       9    0.0     0.000      0.0   0.0      0.864    0.136   \n",
       "332  female  28.0       9    0.0     0.000      0.0   0.0      1.000    0.000   \n",
       "\n",
       "     sadness  surprise  \n",
       "0      0.000       0.0  \n",
       "1      0.000       0.0  \n",
       "2      0.000       0.0  \n",
       "3      0.000       0.0  \n",
       "4      0.000       0.0  \n",
       "..       ...       ...  \n",
       "328    0.000       0.0  \n",
       "329    0.000       0.0  \n",
       "330    0.001       0.0  \n",
       "331    0.000       0.0  \n",
       "332    0.000       0.0  \n",
       "\n",
       "[333 rows x 11 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(data)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Мы загрузили все данные в `DataFrame` по имени `df`, и дальше попробуем его проанализировать. Ниже, когда мы загрузим данные из другого источника, мы сможем перейти снова к этой ячейке и построить графики, просто выполняя ячейки одну за другой.\n",
    "\n",
    "Можем сохранить данные на диск, что впоследствии можно было для демонстрации их быстро загрузить"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_pickle('data_conf.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_pickle('data_conf.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Смотрим на результаты графически\n",
    "\n",
    "Посмотрим на распределение возрастов на конференции:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x0000024FBB72D7C8>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.hist('age')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "А как распределен уровень счастья?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x0000024FBB183148>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.hist('happiness')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Можем также построить зависимость счастья от возраста:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbaed2f48>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.plot.scatter('age','happiness')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим на распределение счастья в зависимости от пола:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbb3b11c8>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.groupby('gender').mean()['happiness'].plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Вывод:** девушки более счастливы!\n",
    "\n",
    "И в зависимости от количества лиц на фотографии:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbb42f088>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.groupby('nfaces').mean()['happiness'].plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Вывод:** фотографироваться в компании немного веселее, но не сильно.\n",
    "\n",
    "Наконец, распределение среднего счасться по возрастным группам:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbb4cf988>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.groupby('age').mean()['happiness'].plot.bar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbb5ef548>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAESCAYAAADtzi4UAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAATYklEQVR4nO3df5BdZ33f8ffHUkx+QCmJN6S1ZNZNxA9Nw5h6EcmkBZraRK4zMi0myDMJkEI0TNGYGTppxKQ1iZl0bNKGZqZKgxNcSKZEcdwJbEBBTgJmknRMtHZciOQRCNXEizEsIEITCEbxt3/cK+9ldXf3rLm7d/e579eMxvec8+js18/V+ejoued5bqoKSdLWd9G4C5AkjYaBLkmNMNAlqREGuiQ1wkCXpEZsH9cPvuSSS2p6enpcP16StqR7773381U1NezY2AJ9enqaubm5cf14SdqSknxquWMOuUhSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiPGNlN0FKYPvX/cJfDgLdeOuwRJArxDl6RmGOiS1AgDXZIaYaBLUiMMdElqhIEuSY0w0CWpEQa6JDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1olOgJ9mb5FSS00kOLdPmx5KcTHIiybtHW6YkaTWrfmNRkm3AYeBqYB44nmS2qk4OtNkFvAn4oao6m+S716tgSdJwXe7Q9wCnq+pMVT0KHAGuW9Lmp4DDVXUWoKo+N9oyJUmr6RLolwIPDWzP9/cNeibwzCR/muSeJHuHnSjJgSRzSeYWFhaeWMWSpKG6BHqG7Ksl29uBXcCLgRuAX0/y9y/4TVW3VdVMVc1MTU2ttVZJ0gq6BPo8sHNgewfw8JA2762qr1fV/wVO0Qt4SdIG6RLox4FdSS5PcjGwH5hd0uY9wD8HSHIJvSGYM6MsVJK0slUDvarOAQeBY8ADwB1VdSLJzUn29ZsdA76Q5CTwIeCnq+oL61W0JOlCqz62CFBVR4GjS/bdNPC6gDf2f0mSxsCZopLUCANdkhphoEtSIwx0SWqEgS5Jjej0lIs2v+lD7x93CTx4y7XjLkGaaN6hS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIa0SnQk+xNcirJ6SSHhhx/dZKFJPf3f7129KVKklay6jcWJdkGHAauBuaB40lmq+rkkqa/XVUH16FGSVIHXe7Q9wCnq+pMVT0KHAGuW9+yJElr1SXQLwUeGtie7+9b6mVJPprkziQ7h50oyYEkc0nmFhYWnkC5kqTldAn0DNlXS7Z/D5iuqucCfwi8a9iJquq2qpqpqpmpqam1VSpJWlGXQJ8HBu+4dwAPDzaoqi9U1df6m78GXDma8iRJXXUJ9OPAriSXJ7kY2A/MDjZI8g8GNvcBD4yuRElSF6s+5VJV55IcBI4B24Dbq+pEkpuBuaqaBW5Msg84B3wRePU61ixJGmLVQAeoqqPA0SX7bhp4/SbgTaMtTZK0Fs4UlaRGGOiS1AgDXZIaYaBLUiMMdElqhIEuSY0w0CWpEQa6JDXCQJekRnSaKSppa5o+9P5xl8CDt1w77hImhoGu5hhimlQOuUhSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiMMdElqhIEuSY3oFOhJ9iY5leR0kkMrtLs+SSWZGV2JkqQuVg30JNuAw8A1wG7ghiS7h7R7CnAj8JFRFylJWl2XO/Q9wOmqOlNVjwJHgOuGtHsL8Fbgb0dYnySpoy6Bfinw0MD2fH/f45I8D9hZVe9b6URJDiSZSzK3sLCw5mIlScvrEugZsq8eP5hcBLwN+HernaiqbquqmaqamZqa6l6lJGlVXQJ9Htg5sL0DeHhg+ynAPwbuTvIg8APArB+MStLG6hLox4FdSS5PcjGwH5g9f7Cq/qqqLqmq6aqaBu4B9lXV3LpULEkaatVAr6pzwEHgGPAAcEdVnUhyc5J9612gJKmbTt8pWlVHgaNL9t20TNsXf/NlSZLWypmiktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiMMdElqhIEuSY0w0CWpEZ0CPcneJKeSnE5yaMjx1yX5WJL7k/xJkt2jL1WStJJVAz3JNuAwcA2wG7hhSGC/u6q+v6quAN4K/NLIK5UkrajLHfoe4HRVnamqR4EjwHWDDarqywOb3wHU6EqUJHWxvUObS4GHBrbngRcsbZTk9cAbgYuBHx52oiQHgAMAl1122VprlSStoMsdeobsu+AOvKoOV9X3Aj8D/IdhJ6qq26pqpqpmpqam1lapJGlFXQJ9Htg5sL0DeHiF9keAl34zRUmS1q5LoB8HdiW5PMnFwH5gdrBBkl0Dm9cCnxhdiZKkLlYdQ6+qc0kOAseAbcDtVXUiyc3AXFXNAgeTXAV8HTgLvGo9i5YkXajLh6JU1VHg6JJ9Nw28fsOI65IkrZEzRSWpEQa6JDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiM6BXqSvUlOJTmd5NCQ429McjLJR5P8UZJnjL5USdJKVg30JNuAw8A1wG7ghiS7lzT7c2Cmqp4L3Am8ddSFSpJW1uUOfQ9wuqrOVNWjwBHgusEGVfWhqvpKf/MeYMdoy5QkraZLoF8KPDSwPd/ft5zXAL8/7ECSA0nmkswtLCx0r1KStKougZ4h+2pow+THgRngF4cdr6rbqmqmqmampqa6VylJWtX2Dm3mgZ0D2zuAh5c2SnIV8LPAi6rqa6MpT5LUVZc79OPAriSXJ7kY2A/MDjZI8jzg7cC+qvrc6MuUJK1m1UCvqnPAQeAY8ABwR1WdSHJzkn39Zr8IPBn4nST3J5ld5nSSpHXSZciFqjoKHF2y76aB11eNuC5J0ho5U1SSGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIzpN/ZekrW760PvHXQIP3nLtup7fO3RJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNaJToCfZm+RUktNJDg05/sIk9yU5l+T60ZcpSVrNqoGeZBtwGLgG2A3ckGT3kmZ/CbwaePeoC5QkddNl+dw9wOmqOgOQ5AhwHXDyfIOqerB/7LF1qFGS1EGXIZdLgYcGtuf7+9YsyYEkc0nmFhYWnsgpJEnL6BLoGbKvnsgPq6rbqmqmqmampqaeyCkkScvoEujzwM6B7R3Aw+tTjiTpieoS6MeBXUkuT3IxsB+YXd+yJElrtWqgV9U54CBwDHgAuKOqTiS5Ock+gCTPTzIPvBx4e5IT61m0JOlCnb4kuqqOAkeX7Ltp4PVxekMxkqQxcaaoJDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiMMdElqRKdAT7I3yakkp5McGnL8SUl+u3/8I0mmR12oJGllqwZ6km3AYeAaYDdwQ5LdS5q9BjhbVd8HvA24ddSFSpJW1uUOfQ9wuqrOVNWjwBHguiVtrgPe1X99J/AvkmR0ZUqSVpOqWrlBcj2wt6pe29/+CeAFVXVwoM1f9NvM97c/2W/z+SXnOgAc6G8+Czg1qv+Rb8IlwOdXbTUZ7Ise+2GRfbFos/TFM6pqatiB7R1+87A77aV/C3RpQ1XdBtzW4WdumCRzVTUz7jo2A/uix35YZF8s2gp90WXIZR7YObC9A3h4uTZJtgNPBb44igIlSd10CfTjwK4klye5GNgPzC5pMwu8qv/6euCDtdpYjiRppFYdcqmqc0kOAseAbcDtVXUiyc3AXFXNAu8AfjPJaXp35vvXs+gR21RDQGNmX/TYD4vsi0Wbvi9W/VBUkrQ1OFNUkhphoEtSIwx0SWqEgS5JjegysUgNSvJUYC9wKb1JYA8Dx6rqS2MtbAz6y1Ts4Rv74s8m8dFb+2LRVrxGJu4pl634Jo1aklcCbwbuAj7d370DuBr4+ar6jXHVttGSvAT4FeATfGNffB/wb6vqrnHVttHsi0Vb9RqZqEDfqm/SqCU5RW+tnS8t2f804CNV9czxVLbxkjwAXFNVDy7ZfzlwtKqeM5bCxsC+WLRVr5FJG3L5WeDK5d4kYCICnd7aO8P+Jn+M4evytGw7vaUrlvo08C0bXMu42ReLtuQ1MmmBviXfpHXwC8B9Se4CHurvu4zev1TeMraqxuN24HiSIyz2xU56s53fMbaqxsO+WLQlr5FJG3J5FXATvSGXC96kqnrnmErbcP1/lfwIvc8SQu/O7FhVnR1rYWOQ5Dn01vQf7IvZqjo51sLGwL5YtBWvkYkKdNiab5IkdTFxz6FX1dmqOlJV/6Wq/nP/tWHel2TTL0C0UZL83Lhr2Czsi0Wb+RqZuEBfzmZ+kzbY28ddwCZy77gL2ETsi0Wb9hqZuCGX5SS5sqr8QytpyzLQJ1B/ctWbgJcC57+b8HPAe4FbJmyS1XbgNcC/Av4hi5PN3gu8o6q+PsbyNpR9sWirXiMTFehb9U0atSTHgA8C76qqR/r7vofet05dVVVXj7O+jZTkt4AvAe9i8RnsHfT64jur6hXjqm2j2ReLtuo1MmmBviXfpFFLcqqqnrXWYy1apS8+vllnBK4H+2LRVr1GJu1D0emquvV8mANU1SNVdSu959EnxaeS/PskTz+/I8nTk/wMi8/nT4qzSV6e5PFrIclFSV4BTNrTT/bFoi15jUxaoG/JN2kdvAL4LuDDSc4m+SJwN/CdwI+Ns7Ax2E/vi80fSfLxJB8HHgH+NVvru3FH4XxffLbfF59gcvtiS14jkzbk8jTgEL2ZcN/d3/1ZYJbeGPrE3IUkeTa98dF7quqvB/bvraoPjK+yjZfkBfQ+APwk8BzgB4CTVXV0rIWNUZLvojfx7r9W1Y+Pu55xS/LP6C0r/LHNvOrkRAX6SpL8ZFX9j3HXsRGS3Ai8HngAuAJ4Q1W9t3/svqr6J+OsbyMleTNwDb11jf6A3kX7YeAqejOIf2GM5W2oJLNDdv8wvc+dqKp9G1vR+CT5s6ra03/9WnrXy3uAlwC/V1W3jLO+5RjofUn+sqomYhw9yceAH6yqv04yDdwJ/GZV/XKSP6+q5421wA3U74srgCfRG17YUVVfTvJt9JZJfe5YC9xASe4DTgK/Tu9fLAF+i/5wS1V9eHzVbazB6yDJceBfVtVCku+g96/a7x9vhcNN1GqLST663CHg6csca9G288MsVfVgkhcDdyZ5BpO16iTAuar6O+ArST5ZVV8GqKqvJnlszLVttBngDfSWmf7pqro/yVcnKcgHXNQfor2I3o3vAkBV/U2Sc+MtbXkTFej0QvtHuPAT+wD/e+PLGZtHklxRVfcD9O/Uf5Te8qmb8s5jHT2a5Nur6ivAled39ucsTFSgV9VjwNuS/E7/v59l8jLivKfSW+4gQCX5nqp6JMmT2cQ3PZP2Zr0PePL5IBuU5O6NL2dsXgl8w11GVZ0DXplk065TsU5eWFVfg8cD7bxvoTc/YeJU1Tzw8iTXAl8edz3jUFXTyxx6jN5M2k3JMXRJasSkPYcuSc0y0CWpEQa6tAbp8brRpuQfTG05Sd6T5N4kJ5Ic6O97TX+6+t1Jfi3Jf+vvn0ryv5Ic7//6oRXOO5XkD5Lcl+TtST6V5JIk00keSPIrwH3AziQ3JPlYkr9IcuvAOQZn3V6f5J391+9M8qtJ/rhf54+uU/doghno2or+TVVdSe+56RuTXAr8R3pT9q8Gnj3Q9peBt1XV84GX0Zs0s5w3Ax/sz5T9Xb5xwbZnAb/Rn2zydeBWerMorwCen+SlHeqeBl4EXAv8apJv7fB7pM4m7bFFteHGJOcfHdsJ/ATw4ar6IkD/OerzS71eBexOHn90+O8leUpV/b8h5/2n9B9Jq6oPJBmcr/Cpqrqn//r5wN3nJ5sk+Z/AC+lNDV/JHf1HIz+R5Ay9v3gueIRWeqIMdG0p/VmtV9FbuuAr/fkDp+gtqjXMRf22X+1y+hWO/U3HdoPPAS+9A1/6jLDPDGukHHLRVvNU4Gw/zJ9Nb5jl24EXJXlael+j9rKB9ncBB89vJLlihXP/Cf2lUZO8BHjaMu0+0v95lyTZBtxAb0Ev6C09+5z+B6dLJ6C8vL+++PcC/4jeX0TSyBjo2mo+AGzvr8vzFuAe4NPAf6IXtH9Ib4Gpv+q3vxGYSfLRJCeB161w7p8HXtJfpOoa4DPABUMzVfUZel9l+CHg/wD3nV+tkt7yzO+jt0LhZ5b81lP0gv/3gddV1d+u4f9bWpUzRdWEJE/ur0mznd4HmrdX1e+u8RxPAv6uqs4l+UHgv1fVSnf0azn3O4H3VdWdozifNIxj6GrFzyW5it649V2s/gHlMJcBd/SHSx4FfmqE9Unrzjt0TZwkP0lvmdhBf1pVrx9HPdKoGOiS1Ag/FJWkRhjoktQIA12SGmGgS1IjDHRJasT/B0BM2TsxAy+iAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['age_group'] = (df['age'] // 10)*10\n",
    "df.groupby('age_group').mean()['happiness'].plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Если счастье убывает, то интересно посмотреть, какие эмоции возрастают. Например, аналогичная динамика для грусти:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbb655bc8>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.groupby('age_group').mean()['sadness'].plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим на средние эмоции людей:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbea37b48>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.mean()[['happiness','sadness','anger','contempt','disgust','fear']].plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Тоже самое по мужчинам и по женщинам:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24fbebab088>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.groupby('gender').mean()[['happiness','sadness','anger','contempt','disgust','fear']].T.plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Анализируем свои фотографии\n",
    "\n",
    "Разместим коллекцию фотографий в директории `images`. Самый простой способ сделать это - загрузить фотографии вручную через интерфейс Azure Notebooks. Для примера я загружу заранее заготовленные фото:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!wget http://www.soshnikov.com/temp/psyimages.zip\n",
    "!unzip psyimages.zip\n",
    "!rm psyimages.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing images\\IMG_0306.jpg 1  faces\n",
      "Processing images\\IMG_0521.jpg 3  faces\n",
      "Processing images\\IMG_0600.jpg 1  faces\n",
      "Processing images\\IMG_0781.jpg 3  faces\n",
      "Processing images\\IMG_0809.jpg 4  faces\n",
      "Processing images\\IMG_0872.jpg 4  faces\n",
      "Processing images\\IMG_0908.jpg 13  faces\n",
      "Processing images\\IMG_0923.jpg 7  faces\n",
      "Processing images\\IMG_0985.jpg 1  faces\n",
      "Processing images\\IMG_1018.jpg 5  faces\n",
      "Processing images\\IMG_1199.jpg 2  faces\n",
      "Processing images\\IMG_1241.jpg 2  faces\n",
      "Processing images\\IMG_1256.jpg 6  faces\n",
      "Found 52 faces\n"
     ]
    }
   ],
   "source": [
    "import PIL, glob, numpy as np\n",
    "\n",
    "data = []\n",
    "for fn in glob.glob(\"images/*\"):\n",
    "    print(\"Processing {}\".format(fn),end=' ')\n",
    "    with open(fn,'rb') as f:\n",
    "        res = cli.face.detect_with_stream(f,return_face_landmarks=False,return_face_id=False,return_face_attributes=['emotion','age','gender'])\n",
    "    nfaces = len(res)\n",
    "    print(nfaces,' faces')\n",
    "    im = PIL.Image.open(fn)\n",
    "    img = np.array(im)\n",
    "    for t in res:\n",
    "        fa = t.as_dict()['face_attributes']\n",
    "        em = fa['emotion']\n",
    "        fr = t.as_dict()['face_rectangle']\n",
    "        face = img[fr['top']:fr['top']+fr['height'],fr['left']:fr['left']+fr['width'],:]\n",
    "        z = { 'gender' : fa['gender'], 'age' : fa['age'], 'nfaces' : nfaces, 'face' : face, 'img' : img  }\n",
    "        data.append(dict(list(z.items())+list(em.items())))\n",
    "\n",
    "print(\"Found {} faces\".format(len(data)))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_full = pd.DataFrame(data)\n",
    "df = df_full.drop(['img','face'],1)\n",
    "df.to_pickle('data_faces.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Мы снова загрузили данные в список `data`, поэтому мы можем перейти вверх к ячейке **Смотрим на результаты графически** и выполнить ячейки, изучая результаты и делая выводы.\n",
    "\n",
    "Поскольку в этом случае у нас также есть сами изображения, попробуем найти самые ярко-выраженные эмоции на фотографиях:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "emotions = ['happiness','sadness','surprise','fear','contempt']\n",
    "emos = { t:(0,None) for t in emotions}\n",
    "for x in data:\n",
    "    for k,v in emos.items():\n",
    "        if x[k]>=v[0]:\n",
    "            emos[k]=(x[k],x['face'])\n",
    "fix,ax = plt.subplots(1,len(emos.keys()))\n",
    "for i,e in enumerate(emos.keys()):\n",
    "    if emos[e][1] is not None:\n",
    "        ax[i].imshow(emos[e][1])\n",
    "        ax[i].axis('off')\n",
    "        ax[i].set_title(e)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Теперь найдём наиболее \"противоречивую\" эмоцию, в которой две составляющих проявлены наиболее сильно. Для этого ищем лицо, в котором максимально значение второй по величине эмоции, не считая нейтральной:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'happiness': 0.087,\n",
       " 'sadness': 0.0,\n",
       " 'surprise': 0.057,\n",
       " 'fear': 0.0,\n",
       " 'contempt': 0.0}"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mx = 0\n",
    "img = None\n",
    "emo = None\n",
    "for x in data:\n",
    "    em = { k: x[k] for k in emotions }\n",
    "    snd = sorted(em.values())[-2]\n",
    "    if snd>mx:\n",
    "        mx = snd\n",
    "        img = x['face']\n",
    "        emo = em\n",
    "plt.imshow(img)\n",
    "plt.show()\n",
    "emo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Дальнейшая обработка результатов методами Data Science\n",
    "\n",
    "Полученные результаты могут использоваться для дальнейших исследований с помощью алгоритмов машинного обучения. Например, мы увидели выше интресное сочетание эмоций. Можем посмотреть, выделяются ли среди всех лиц некоторые \"типичные\" конфигурации. Для этого необходимо применить алгоритм кластеризации, который реализован в библиотеке *Scikit Learn*:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sklearn.cluster\n",
    "inp = np.array([ [ x[e] for e in emotions ] for x in data])\n",
    "centroid, labels, inertia = sklearn.cluster.k_means(inp,3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Мы попросили разбить все лица на 3 кластера, на основе расстояний между векторами эмоций. Теперь посмотрим на то, какие конфигурации эмоций у нас получились в центре таких кластеров (они возвращены в переменной `centroids`), и на наиболее \"характерные\" (близкие по расстоянию) лица для каждого из кластеров"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "def closest_face(e):\n",
    "    mi = 999\n",
    "    el = None\n",
    "    for i,x in enumerate(data):\n",
    "        if np.linalg.norm(inp[i]-e)<mi:\n",
    "            mi=np.linalg.norm(inp[i]-e)\n",
    "            el=x\n",
    "    return el['face']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,ax = plt.subplots(2,len(centroid),figsize=(10,4))\n",
    "for i,x in enumerate(centroid):\n",
    "    ax[0,i].bar([t[0:2] for t in emotions],x)\n",
    "    ax[1,i].imshow(closest_face(x))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Посмотрим на все лица в интересном кластере 0:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 17 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = [x['face'] for x,i in zip(data,labels) if i==0]\n",
    "fig,ax = plt.subplots(1,len(res),figsize=(20,4))\n",
    "for i,x in enumerate(res):\n",
    "    ax[i].imshow(x)\n",
    "    ax[i].axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Экспериментируйте!\n",
    "\n",
    "Не забывайте, что вы можете всегда загрузить новые фотографии в директорию `images`, и повторить данный эксперимент для новых данных. Более того, вы можете использовать этот ноутбук как заготовку для дальнейших исследований и экспериментов! Если у вас будет получаться что-то интересное --- делитесь результатами!\n",
    "\n",
    "Контакты:\n",
    "* Дмитрий Сошников\n",
    "* Cloud Developer Advocate, Microsoft\n",
    "* Доцент, МФТИ/НИУ ВШЭ/МАИ\n",
    "* http://soshnikov.com\n",
    "* http://t.me/shwars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "!rm images/*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}