quick_start.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Pdt8PbiPK7hq"
      },
      "source": [
        "# Fake data generator"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "cP2oIKn6CK9G",
        "outputId": "ac374882-8b05-4bf3-ae55-31487b836316"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Collecting aitextgen\n",
            "  Downloading aitextgen-0.5.2.tar.gz (572 kB)\n",
            "\u001b[K     |████████████████████████████████| 572 kB 5.1 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting transformers>=4.5.1\n",
            "  Downloading transformers-4.17.0-py3-none-any.whl (3.8 MB)\n",
            "\u001b[K     |████████████████████████████████| 3.8 MB 3.6 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting fire>=0.3.0\n",
            "  Downloading fire-0.4.0.tar.gz (87 kB)\n",
            "\u001b[K     |████████████████████████████████| 87 kB 8.1 MB/s  eta 0:00:01\n",
            "\u001b[?25hCollecting pytorch-lightning>=1.3.1\n",
            "  Downloading pytorch_lightning-1.5.10-py3-none-any.whl (527 kB)\n",
            "\u001b[K     |████████████████████████████████| 527 kB 2.4 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting torch>=1.6.0\n",
            "  Downloading torch-1.11.0-cp38-none-macosx_10_9_x86_64.whl (129.9 MB)\n",
            "\u001b[K     |████████████████████████████████| 129.9 MB 864 kB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: six in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from fire>=0.3.0->aitextgen) (1.15.0)\n",
            "Collecting termcolor\n",
            "  Downloading termcolor-1.1.0.tar.gz (3.9 kB)\n",
            "Requirement already satisfied: future>=0.17.1 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from pytorch-lightning>=1.3.1->aitextgen) (0.18.2)\n",
            "Requirement already satisfied: PyYAML>=5.1 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from pytorch-lightning>=1.3.1->aitextgen) (5.4.1)\n",
            "Collecting torchmetrics>=0.4.1\n",
            "  Downloading torchmetrics-0.7.2-py3-none-any.whl (397 kB)\n",
            "\u001b[K     |████████████████████████████████| 397 kB 4.9 MB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.17.2 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from pytorch-lightning>=1.3.1->aitextgen) (1.20.3)\n",
            "Collecting typing-extensions\n",
            "  Downloading typing_extensions-4.1.1-py3-none-any.whl (26 kB)\n",
            "Collecting setuptools==59.5.0\n",
            "  Downloading setuptools-59.5.0-py3-none-any.whl (952 kB)\n",
            "\u001b[K     |████████████████████████████████| 952 kB 364 kB/s eta 0:00:01\n",
            "\u001b[?25hCollecting pyDeprecate==0.3.1\n",
            "  Downloading pyDeprecate-0.3.1-py3-none-any.whl (10 kB)\n",
            "Requirement already satisfied: tqdm>=4.41.0 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from pytorch-lightning>=1.3.1->aitextgen) (4.60.0)\n",
            "Collecting tensorboard>=2.2.0\n",
            "  Downloading tensorboard-2.8.0-py3-none-any.whl (5.8 MB)\n",
            "\u001b[K     |████████████████████████████████| 5.8 MB 1.9 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting fsspec[http]!=2021.06.0,>=2021.05.0\n",
            "  Downloading fsspec-2022.2.0-py3-none-any.whl (134 kB)\n",
            "\u001b[K     |████████████████████████████████| 134 kB 2.5 MB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: packaging>=17.0 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from pytorch-lightning>=1.3.1->aitextgen) (21.0)\n",
            "Requirement already satisfied: requests in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from fsspec[http]!=2021.06.0,>=2021.05.0->pytorch-lightning>=1.3.1->aitextgen) (2.25.1)\n",
            "Collecting aiohttp\n",
            "  Downloading aiohttp-3.8.1-cp38-cp38-macosx_10_9_x86_64.whl (574 kB)\n",
            "\u001b[K     |████████████████████████████████| 574 kB 4.9 MB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: pyparsing>=2.0.2 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from packaging>=17.0->pytorch-lightning>=1.3.1->aitextgen) (2.4.7)\n",
            "Collecting grpcio>=1.24.3\n",
            "  Downloading grpcio-1.44.0-cp38-cp38-macosx_10_10_x86_64.whl (4.3 MB)\n",
            "\u001b[K     |████████████████████████████████| 4.3 MB 5.1 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting absl-py>=0.4\n",
            "  Downloading absl_py-1.0.0-py3-none-any.whl (126 kB)\n",
            "\u001b[K     |████████████████████████████████| 126 kB 2.6 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting werkzeug>=0.11.15\n",
            "  Downloading Werkzeug-2.0.3-py3-none-any.whl (289 kB)\n",
            "\u001b[K     |████████████████████████████████| 289 kB 2.7 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting google-auth-oauthlib<0.5,>=0.4.1\n",
            "  Downloading google_auth_oauthlib-0.4.6-py2.py3-none-any.whl (18 kB)\n",
            "Requirement already satisfied: protobuf>=3.6.0 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from tensorboard>=2.2.0->pytorch-lightning>=1.3.1->aitextgen) (3.19.3)\n",
            "Collecting tensorboard-data-server<0.7.0,>=0.6.0\n",
            "  Downloading tensorboard_data_server-0.6.1-py3-none-macosx_10_9_x86_64.whl (3.5 MB)\n",
            "\u001b[K     |████████████████████████████████| 3.5 MB 402 kB/s eta 0:00:01\n",
            "\u001b[?25hCollecting google-auth<3,>=1.6.3\n",
            "  Downloading google_auth-2.6.2-py2.py3-none-any.whl (156 kB)\n",
            "\u001b[K     |████████████████████████████████| 156 kB 5.4 MB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: wheel>=0.26 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from tensorboard>=2.2.0->pytorch-lightning>=1.3.1->aitextgen) (0.36.2)\n",
            "Collecting tensorboard-plugin-wit>=1.6.0\n",
            "  Downloading tensorboard_plugin_wit-1.8.1-py3-none-any.whl (781 kB)\n",
            "\u001b[K     |████████████████████████████████| 781 kB 4.0 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting markdown>=2.6.8\n",
            "  Downloading Markdown-3.3.6-py3-none-any.whl (97 kB)\n",
            "\u001b[K     |████████████████████████████████| 97 kB 799 kB/s eta 0:00:011\n",
            "\u001b[?25hCollecting pyasn1-modules>=0.2.1\n",
            "  Downloading pyasn1_modules-0.2.8-py2.py3-none-any.whl (155 kB)\n",
            "\u001b[K     |████████████████████████████████| 155 kB 3.1 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting cachetools<6.0,>=2.0.0\n",
            "  Downloading cachetools-5.0.0-py3-none-any.whl (9.1 kB)\n",
            "Collecting rsa<5,>=3.1.4\n",
            "  Downloading rsa-4.8-py3-none-any.whl (39 kB)\n",
            "Collecting requests-oauthlib>=0.7.0\n",
            "  Downloading requests_oauthlib-1.3.1-py2.py3-none-any.whl (23 kB)\n",
            "Requirement already satisfied: importlib-metadata>=4.4 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from markdown>=2.6.8->tensorboard>=2.2.0->pytorch-lightning>=1.3.1->aitextgen) (4.6.4)\n",
            "Requirement already satisfied: zipp>=0.5 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from importlib-metadata>=4.4->markdown>=2.6.8->tensorboard>=2.2.0->pytorch-lightning>=1.3.1->aitextgen) (3.5.0)\n",
            "Collecting pyasn1<0.5.0,>=0.4.6\n",
            "  Downloading pyasn1-0.4.8-py2.py3-none-any.whl (77 kB)\n",
            "\u001b[K     |████████████████████████████████| 77 kB 2.3 MB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: chardet<5,>=3.0.2 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from requests->fsspec[http]!=2021.06.0,>=2021.05.0->pytorch-lightning>=1.3.1->aitextgen) (4.0.0)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from requests->fsspec[http]!=2021.06.0,>=2021.05.0->pytorch-lightning>=1.3.1->aitextgen) (2021.5.30)\n",
            "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from requests->fsspec[http]!=2021.06.0,>=2021.05.0->pytorch-lightning>=1.3.1->aitextgen) (1.26.4)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from requests->fsspec[http]!=2021.06.0,>=2021.05.0->pytorch-lightning>=1.3.1->aitextgen) (2.10)\n",
            "Collecting oauthlib>=3.0.0\n",
            "  Downloading oauthlib-3.2.0-py3-none-any.whl (151 kB)\n",
            "\u001b[K     |████████████████████████████████| 151 kB 9.7 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting tokenizers!=0.11.3,>=0.11.1\n",
            "  Downloading tokenizers-0.11.6-cp38-cp38-macosx_10_11_x86_64.whl (3.6 MB)\n",
            "\u001b[K     |████████████████████████████████| 3.6 MB 1.1 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting huggingface-hub<1.0,>=0.1.0\n",
            "  Downloading huggingface_hub-0.4.0-py3-none-any.whl (67 kB)\n",
            "\u001b[K     |████████████████████████████████| 67 kB 3.1 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting filelock\n",
            "  Downloading filelock-3.6.0-py3-none-any.whl (10.0 kB)\n",
            "Collecting sacremoses\n",
            "  Downloading sacremoses-0.0.49-py3-none-any.whl (895 kB)\n",
            "\u001b[K     |████████████████████████████████| 895 kB 3.3 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting regex!=2019.12.17\n",
            "  Downloading regex-2022.3.15-cp38-cp38-macosx_10_9_x86_64.whl (289 kB)\n",
            "\u001b[K     |████████████████████████████████| 289 kB 6.9 MB/s eta 0:00:01\n",
            "\u001b[?25hCollecting aiosignal>=1.1.2\n",
            "  Downloading aiosignal-1.2.0-py3-none-any.whl (8.2 kB)\n",
            "Collecting yarl<2.0,>=1.0\n",
            "  Downloading yarl-1.7.2-cp38-cp38-macosx_10_9_x86_64.whl (121 kB)\n",
            "\u001b[K     |████████████████████████████████| 121 kB 6.4 MB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: attrs>=17.3.0 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from aiohttp->fsspec[http]!=2021.06.0,>=2021.05.0->pytorch-lightning>=1.3.1->aitextgen) (21.2.0)\n",
            "Collecting async-timeout<5.0,>=4.0.0a3\n",
            "  Downloading async_timeout-4.0.2-py3-none-any.whl (5.8 kB)\n",
            "Collecting frozenlist>=1.1.1\n",
            "  Downloading frozenlist-1.3.0-cp38-cp38-macosx_10_9_x86_64.whl (36 kB)\n",
            "Collecting charset-normalizer<3.0,>=2.0\n",
            "  Downloading charset_normalizer-2.0.12-py3-none-any.whl (39 kB)\n",
            "Collecting multidict<7.0,>=4.5\n",
            "  Downloading multidict-6.0.2-cp38-cp38-macosx_10_9_x86_64.whl (28 kB)\n",
            "Requirement already satisfied: joblib in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from sacremoses->transformers>=4.5.1->aitextgen) (1.0.1)\n",
            "Requirement already satisfied: click in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from sacremoses->transformers>=4.5.1->aitextgen) (7.1.2)\n",
            "Building wheels for collected packages: aitextgen, fire, termcolor\n",
            "  Building wheel for aitextgen (setup.py) ... \u001b[?25ldone\n",
            "\u001b[?25h  Created wheel for aitextgen: filename=aitextgen-0.5.2-py3-none-any.whl size=575888 sha256=74705025c1840e4696cd3bd284c8afdebc7cb60ed3123a4caa475f259aa6f49b\n",
            "  Stored in directory: /Users/liangsiyun/Library/Caches/pip/wheels/78/25/a3/ae50899e988bc7b31b27944ef406c5d460e99336255053e871\n",
            "  Building wheel for fire (setup.py) ... \u001b[?25ldone\n",
            "\u001b[?25h  Created wheel for fire: filename=fire-0.4.0-py2.py3-none-any.whl size=115928 sha256=ae3b2ac222062829e8b8b542fed91714421fb13ec7782a69603c47befc6fd2d5\n",
            "  Stored in directory: /Users/liangsiyun/Library/Caches/pip/wheels/1f/10/06/2a990ee4d73a8479fe2922445e8a876d38cfbfed052284c6a1\n",
            "  Building wheel for termcolor (setup.py) ... \u001b[?25ldone\n",
            "\u001b[?25h  Created wheel for termcolor: filename=termcolor-1.1.0-py3-none-any.whl size=4829 sha256=e382513b59f630edb73a4f30d44a77d3f608700737dd05320a0be189a045feda\n",
            "  Stored in directory: /Users/liangsiyun/Library/Caches/pip/wheels/a0/16/9c/5473df82468f958445479c59e784896fa24f4a5fc024b0f501\n",
            "Successfully built aitextgen fire termcolor\n",
            "Installing collected packages: pyasn1, rsa, pyasn1-modules, oauthlib, multidict, frozenlist, cachetools, yarl, typing-extensions, requests-oauthlib, google-auth, charset-normalizer, async-timeout, aiosignal, werkzeug, torch, tensorboard-plugin-wit, tensorboard-data-server, setuptools, regex, pyDeprecate, markdown, grpcio, google-auth-oauthlib, fsspec, filelock, aiohttp, absl-py, torchmetrics, tokenizers, termcolor, tensorboard, sacremoses, huggingface-hub, transformers, pytorch-lightning, fire, aitextgen\n",
            "  Attempting uninstall: setuptools\n",
            "    Found existing installation: setuptools 52.0.0.post20210125\n",
            "    Uninstalling setuptools-52.0.0.post20210125:\n",
            "      Successfully uninstalled setuptools-52.0.0.post20210125\n",
            "  Attempting uninstall: fsspec\n",
            "    Found existing installation: fsspec 2021.6.0\n",
            "    Uninstalling fsspec-2021.6.0:\n",
            "      Successfully uninstalled fsspec-2021.6.0\n",
            "Successfully installed absl-py-1.0.0 aiohttp-3.8.1 aiosignal-1.2.0 aitextgen-0.5.2 async-timeout-4.0.2 cachetools-5.0.0 charset-normalizer-2.0.12 filelock-3.6.0 fire-0.4.0 frozenlist-1.3.0 fsspec-2022.2.0 google-auth-2.6.2 google-auth-oauthlib-0.4.6 grpcio-1.44.0 huggingface-hub-0.4.0 markdown-3.3.6 multidict-6.0.2 oauthlib-3.2.0 pyDeprecate-0.3.1 pyasn1-0.4.8 pyasn1-modules-0.2.8 pytorch-lightning-1.5.10 regex-2022.3.15 requests-oauthlib-1.3.1 rsa-4.8 sacremoses-0.0.49 setuptools-59.5.0 tensorboard-2.8.0 tensorboard-data-server-0.6.1 tensorboard-plugin-wit-1.8.1 termcolor-1.1.0 tokenizers-0.11.6 torch-1.11.0 torchmetrics-0.7.2 transformers-4.17.0 typing-extensions-4.1.1 werkzeug-2.0.3 yarl-1.7.2\n",
            "Note: you may need to restart the kernel to use updated packages.\n"
          ]
        }
      ],
      "source": [
        "pip install aitextgen"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8p-MK4fHCK9G",
        "outputId": "4060a12a-2a7b-4bd4-8b76-28dc5827400a"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Collecting Faker\n",
            "  Downloading Faker-13.3.2-py3-none-any.whl (1.5 MB)\n",
            "\u001b[K     |████████████████████████████████| 1.5 MB 4.8 MB/s eta 0:00:01\n",
            "\u001b[?25hRequirement already satisfied: python-dateutil>=2.4 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from Faker) (2.8.1)\n",
            "Requirement already satisfied: six>=1.5 in /opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages (from python-dateutil>=2.4->Faker) (1.15.0)\n",
            "Installing collected packages: Faker\n",
            "Successfully installed Faker-13.3.2\n",
            "Note: you may need to restart the kernel to use updated packages.\n"
          ]
        }
      ],
      "source": [
        "pip install Faker"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9Adgf-UIK7hw"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "from aitextgen.TokenDataset import TokenDataset\n",
        "from aitextgen.tokenizers import train_tokenizer\n",
        "from aitextgen.utils import GPT2ConfigCPU\n",
        "from aitextgen import aitextgen\n",
        "from faker import Faker\n",
        "from random import shuffle, sample, randint, random\n",
        "import ast\n",
        "import re\n",
        "\n",
        "\n",
        "class artificial_data(object):\n",
        "\n",
        "    def __init__(self):\n",
        "        # Train a data2text model to generate artificial description of project on the website\n",
        "        file_name = 'dataset_des.txt'\n",
        "        self.ai_text_generator(file_name)\n",
        "        # Load the trained model\n",
        "        self.ai_text = aitextgen(model_folder=\"trained_model\", tokenizer_file=\"aitextgen.tokenizer.json\")\n",
        "\n",
        "        # Use Faker package to generate fake users' information\n",
        "        self.fake = Faker()\n",
        "\n",
        "        # Mask users' features\n",
        "        all_user_features = pd.read_csv('user_feature.csv',index_col=0)\n",
        "        all_user_features = self.user_feature_mask(all_user_features)\n",
        "        all_user_features = self.clean_all_user_features(all_user_features)\n",
        "        all_user_features.to_csv('fake_all_user_features.csv')\n",
        "\n",
        "        # Mask users' activities features\n",
        "        # Join the users' activities datasets into one dataframe as match_df\n",
        "        datasets_list = ['top_click_all.csv','visits_30_user.csv','course_enrolments.csv','order_info.csv']\n",
        "        match_df = self.concat_match_df(datasets_list,all_user_features)\n",
        "        match_df = self.match_df_mask(match_df)\n",
        "        match_df.to_csv('match_df.csv')\n",
        "\n",
        "\n",
        "\n",
        "    def ai_text_generator(self,file_name):\n",
        "        # Train a custom BPE Tokenizer on the downloaded text\n",
        "        # This will save one file: `aitextgen.tokenizer.json`, which contains the\n",
        "        # information needed to rebuild the tokenizer.\n",
        "        train_tokenizer(file_name)\n",
        "        tokenizer_file = \"aitextgen.tokenizer.json\"\n",
        "\n",
        "        # GPT2ConfigCPU is a mini variant of GPT-2 optimized for CPU-training\n",
        "        config = GPT2ConfigCPU()\n",
        "\n",
        "        # Instantiate aitextgen using the created tokenizer and config\n",
        "        ai = aitextgen(tokenizer_file=tokenizer_file, config=config)\n",
        "\n",
        "        # Build datasets for training by creating TokenDatasets,\n",
        "        # which automatically processes the dataset with the appropriate size.\n",
        "        data = TokenDataset(file_name, tokenizer_file=tokenizer_file, block_size=64)\n",
        "\n",
        "        # Train the model. It will save pytorch_model.bin periodically and after completion to the `trained_model` folder.\n",
        "        # On a 2019 Macbook Pro, this took ~50 minutes to run.\n",
        "        ai.train(data, batch_size=8, num_steps=50000, generate_every=5000, save_every=5000)\n",
        "        return None\n",
        "\n",
        "\n",
        "\n",
        "    def user_feature_mask(self,all_user_features):\n",
        "        # fake username of the email address\n",
        "        all_user_features['email'] = [self.fake.user_name()+ '@' + all_user_features['email'].values[i].split('@')[1]\n",
        "                                    for i in range(0,all_user_features.shape[0])]\n",
        "\n",
        "        # fake username, first_name, last_name\n",
        "        all_user_features['username'] = [self.fake.user_name() for _ in range(0,all_user_features.shape[0])]\n",
        "        all_user_features['first_name'] = [self.fake.first_name() for _ in range(0,all_user_features.shape[0])]\n",
        "        all_user_features['last_name'] = [self.fake.last_name() for _ in range(0,all_user_features.shape[0])]\n",
        "        \n",
        "        # shuffle and resample company, job_title, education, university, major\n",
        "        all_user_features['company'] = self.shuffle_list(list(all_user_features['company']))\n",
        "        all_user_features['job_title'] = sample(list(all_user_features['job_title'])+\n",
        "                                        list(set(all_user_features['job_title']))+\n",
        "                                        [self.fake.job() for _ in range(40)],k=all_user_features.shape[0])\n",
        "        all_user_features['education'] = sample(list(all_user_features['education'])+\n",
        "                                        list(set(all_user_features['education']))*20,\n",
        "                                        k=all_user_features.shape[0])\n",
        "        all_user_features['university'] = sample(list(all_user_features['university'])+\n",
        "                                        list(set(all_user_features['university']))*5,\n",
        "                                        k=all_user_features.shape[0])\n",
        "        all_user_features['major'] = sample(list(all_user_features['major'])+\n",
        "                                        list(set(all_user_features['major']))*5,\n",
        "                                        k=all_user_features.shape[0])\n",
        "        \n",
        "        # shuffle cols 11 - 40, 42 - 47\n",
        "        for i in list(range(11,41))+list(range(42,48)):\n",
        "            all_user_features.iloc[:,i] = self.shuffle_list(list(all_user_features.iloc[:,i]))\n",
        "\n",
        "        # project_list, randomly append 200 project information in the null rows\n",
        "        notnull_row_origin = list(all_user_features.loc[all_user_features['project_list'].notnull()].index)\n",
        "        self.random_project_list(all_user_features,200)\n",
        "        \n",
        "        # num_follower, append random num to those users who is appended random project_list on previous step\n",
        "        notnull_row = list(all_user_features.loc[all_user_features['project_list'].notnull()].index)\n",
        "        append_row = list(set(notnull_row)-set(notnull_row_origin))\n",
        "        all_user_features.loc[append_row,'num_follower'] = [randint(1,50) for _ in range(len(append_row))]\n",
        "\n",
        "        # impact score, append impact score to those users who is appended random project_list on previous step\n",
        "        # impact score range: 0 - 10\n",
        "        all_user_features.loc[append_row,'impact_score'] = [random()*10 for _ in range(len(append_row))]\n",
        "        \n",
        "        # Turn string into list syntax\n",
        "        all_user_features[['project_list']] = all_user_features[['project_list']].applymap(self.literal_return)\n",
        "\n",
        "        return all_user_features\n",
        "\n",
        "\n",
        "    def shuffle_list(self,df_col):\n",
        "        shuffle(df_col)\n",
        "        return df_col\n",
        "\n",
        "    def random_project_list(self,all_user_features,num):\n",
        "        # Random num of project in each list (range: 1-10)\n",
        "        # project id is random number from 0 to 2000\n",
        "        # project title and description generated by ai text generater, title with max_length 15\n",
        "        # description with max_length 150\n",
        "        # programing language is randomly chosen from a list of programing software\n",
        "        programing = ['Python','R','C++','C','Other','Java',None]\n",
        "        null_row = list(all_user_features.loc[all_user_features['project_list'].isnull()].index)\n",
        "        random_append_index = sample(null_row,k = num)\n",
        "        for i in random_append_index:\n",
        "            project_num = randint(1,10)\n",
        "            project_list = []\n",
        "            for _ in range(project_num):\n",
        "                info = (randint(0,2000),self.ai_text.generate(1,max_length = 15,return_as_list=True)[0],\n",
        "                        self.ai_text.generate(1,max_length = 150,return_as_list=True)[0],\n",
        "                        sample(programing,k=1)[0])\n",
        "                project_list.append(info)\n",
        "            all_user_features.loc[i,'project_list'] = str(project_list)\n",
        "        return None\n",
        "\n",
        "    def clean_all_user_features(self,all_user_features):\n",
        "        # Clean the text data and turn them into lower case\n",
        "        clean_list = ['company','job_title','education','major','university']\n",
        "        for l in clean_list:\n",
        "            all_user_features[l] = all_user_features[l].apply(lambda x: re.sub(r'[^\\w\\s]', '', str(x)).lower())\n",
        "        return all_user_features\n",
        "\n",
        "    def literal_return(self,val):\n",
        "        # Turn the string into list syntax\n",
        "        try:\n",
        "            return ast.literal_eval(val)\n",
        "        except (ValueError, SyntaxError) as e:\n",
        "            return None\n",
        "\n",
        "\n",
        "\n",
        "    def concat_match_df(self,datasets_list,all_user_features):\n",
        "        match_df = pd.DataFrame()\n",
        "        for file in datasets_list:\n",
        "            df = pd.read_csv(file, index_col=0)\n",
        "            match_df = pd.concat([match_df,df],axis=1)\n",
        "\n",
        "        # Select uid corresponding with all_user_features\n",
        "        match_df = match_df.loc[list(set(all_user_features.index)&set(match_df.index)),:]\n",
        "\n",
        "        # Clean the text data\n",
        "        clean_list = ['all_enrol','all_visits','all_orders']\n",
        "        for l in clean_list:\n",
        "            match_df[[l]] = match_df[[l]].applymap(self.str_to_list)\n",
        "        return match_df\n",
        "\n",
        "    def str_to_list(self,string):\n",
        "        # Turn the string into list syntax\n",
        "        if str(string) not in ['NaN','nan','None']:\n",
        "            return string.strip('][').split(', ')\n",
        "        else:\n",
        "            return None\n",
        "\n",
        "    def match_df_mask(self,match_df):\n",
        "        # expand purchase list\n",
        "        order_num = [randint(200,300) for _ in range(50)]\n",
        "        shuffle_order = list(set(sample(list(match_df[match_df.all_orders.isnull()].index),k=180)))\n",
        "        match_df.loc[shuffle_order,'all_orders'] = sample(order_num*5,k=len(shuffle_order))\n",
        "        match_df.loc[shuffle_order,'purchased_entity'] = match_df.loc[shuffle_order,'all_orders']\n",
        "        # expand visits list\n",
        "        shuffle_visit = list(set(sample(list(match_df[match_df.all_visits.isnull()].index),k=320)))\n",
        "        visit_info = sample(list(match_df[match_df.all_visits.notnull()].index)*10,k=len(shuffle_visit))\n",
        "        for col in ['url','time','time_period','all_visits']:\n",
        "            match_df.loc[shuffle_visit,col] = list(match_df.loc[visit_info,col])\n",
        "\n",
        "        # expand enrollment\n",
        "        shuffle_enrol = list(set(sample(list(match_df[match_df.all_enrol.isnull()].index),k=170)))\n",
        "        enrol_info = list(set(match_df[match_df.all_enrol.notnull()]['enrolment']))\n",
        "        match_df.loc[shuffle_enrol,'enrolment'] = sample(enrol_info*10,k=len(shuffle_enrol))\n",
        "        match_df.loc[shuffle_enrol,'all_enrol'] = match_df.loc[shuffle_enrol,'enrolment']\n",
        "        return match_df\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "LP29oGVvK7h1",
        "outputId": "23f98561-edc7-43af-d065-5cdfcc21af9a"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages/numpy/core/_asarray.py:102: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
            "  return array(a, dtype, copy=False, order=order)\n",
            "/opt/anaconda3/envs/rqsdk/lib/python3.8/site-packages/pandas/core/internals/blocks.py:867: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
            "  arr_value = np.array(value)\n"
          ]
        }
      ],
      "source": [
        "fake_data = artificial_data()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 62,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "1wOX-AFMK7h2",
        "outputId": "fc9456f1-58d6-428d-9462-f476e5cbeb28"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 884 entries, 0 to 883\n",
            "Data columns (total 49 columns):\n",
            " #   Column                        Non-Null Count  Dtype  \n",
            "---  ------                        --------------  -----  \n",
            " 0   uid                           884 non-null    int64  \n",
            " 1   email                         884 non-null    object \n",
            " 2   username                      884 non-null    object \n",
            " 3   first_name                    884 non-null    object \n",
            " 4   last_name                     884 non-null    object \n",
            " 5   impact_score                  884 non-null    float64\n",
            " 6   company                       487 non-null    object \n",
            " 7   job_title                     488 non-null    object \n",
            " 8   education                     131 non-null    object \n",
            " 9   university                    172 non-null    object \n",
            " 10  major                         160 non-null    object \n",
            " 11  num_follower                  884 non-null    float64\n",
            " 12  communication_services        884 non-null    float64\n",
            " 13  consumer_discretionary        884 non-null    float64\n",
            " 14  consumer_staples              884 non-null    float64\n",
            " 15  energy                        884 non-null    float64\n",
            " 16  financials                    884 non-null    float64\n",
            " 17  health_care                   884 non-null    float64\n",
            " 18  industrials                   884 non-null    float64\n",
            " 19  information_technology        884 non-null    float64\n",
            " 20  materials                     884 non-null    float64\n",
            " 21  real_estate                   884 non-null    float64\n",
            " 22  utilities                     884 non-null    float64\n",
            " 23  Fraud_Detection               884 non-null    float64\n",
            " 24  Risk_Scoring                  884 non-null    float64\n",
            " 25  Healthcare                    884 non-null    float64\n",
            " 26  Internet_Search               884 non-null    float64\n",
            " 27  Marketing_Effectiveness       884 non-null    float64\n",
            " 28  Website_Recommendations       884 non-null    float64\n",
            " 29  Image_Recognition             884 non-null    float64\n",
            " 30  Speech_Recognition            884 non-null    float64\n",
            " 31  Airline_Route_Planning        884 non-null    float64\n",
            " 32  Price_Analytics               884 non-null    float64\n",
            " 33  Supply_Chain_Optimization     884 non-null    float64\n",
            " 34  Talent_Acquisition_Analytics  884 non-null    float64\n",
            " 35  Environment_Analytics         884 non-null    float64\n",
            " 36  Epidemiology                  884 non-null    float64\n",
            " 37  Social_Policy                 884 non-null    float64\n",
            " 38  Evaluation_and_Assessment     884 non-null    float64\n",
            " 39  filter_test                   884 non-null    bool   \n",
            " 40  filter_spam                   884 non-null    bool   \n",
            " 41  filter_spam_2                 884 non-null    bool   \n",
            " 42  project_list                  234 non-null    object \n",
            " 43  covid_competition             884 non-null    bool   \n",
            " 44  news_competition              884 non-null    bool   \n",
            " 45  property_competition          884 non-null    bool   \n",
            " 46  restaurant_competition        884 non-null    bool   \n",
            " 47  investment_competition        884 non-null    bool   \n",
            " 48  recency                       884 non-null    int64  \n",
            "dtypes: bool(8), float64(29), int64(2), object(10)\n",
            "memory usage: 290.2+ KB\n",
            "None\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-68f14015-8a28-4c19-9505-b008879a1837\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>uid</th>\n",
              "      <th>email</th>\n",
              "      <th>username</th>\n",
              "      <th>first_name</th>\n",
              "      <th>last_name</th>\n",
              "      <th>impact_score</th>\n",
              "      <th>company</th>\n",
              "      <th>job_title</th>\n",
              "      <th>education</th>\n",
              "      <th>university</th>\n",
              "      <th>...</th>\n",
              "      <th>filter_test</th>\n",
              "      <th>filter_spam</th>\n",
              "      <th>filter_spam_2</th>\n",
              "      <th>project_list</th>\n",
              "      <th>covid_competition</th>\n",
              "      <th>news_competition</th>\n",
              "      <th>property_competition</th>\n",
              "      <th>restaurant_competition</th>\n",
              "      <th>investment_competition</th>\n",
              "      <th>recency</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>678</td>\n",
              "      <td>christopherknight@rmdslab.com</td>\n",
              "      <td>castrochristina</td>\n",
              "      <td>Christopher</td>\n",
              "      <td>Long</td>\n",
              "      <td>40.460932</td>\n",
              "      <td>university of st thomas</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>[(489, 'RMDS Meetup Speakers Presentations', N...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>246</td>\n",
              "      <td>cgreene@gmail.com</td>\n",
              "      <td>gtucker</td>\n",
              "      <td>Jennifer</td>\n",
              "      <td>Robbins</td>\n",
              "      <td>29.369300</td>\n",
              "      <td>simon fraser university</td>\n",
              "      <td>ceo</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>[(376, 'Linear Regression Analysis of US Censu...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2232</td>\n",
              "      <td>hthomas@gmail.com</td>\n",
              "      <td>zrodriguez</td>\n",
              "      <td>Spencer</td>\n",
              "      <td>Ramos</td>\n",
              "      <td>23.920091</td>\n",
              "      <td>universitas pasundan</td>\n",
              "      <td>theatre stage manager</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>[(884, 'Bank Churn', None, None), (1251, 'Educ...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3252</td>\n",
              "      <td>kimjanice@gmail.com</td>\n",
              "      <td>lucashall</td>\n",
              "      <td>Larry</td>\n",
              "      <td>Shah</td>\n",
              "      <td>15.879526</td>\n",
              "      <td>NaN</td>\n",
              "      <td>student</td>\n",
              "      <td>NaN</td>\n",
              "      <td>adler university</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>9174</td>\n",
              "      <td>haroldguerrero@gmail.com</td>\n",
              "      <td>debbiekelley</td>\n",
              "      <td>Craig</td>\n",
              "      <td>Tanner</td>\n",
              "      <td>14.205337</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>less than high school</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>[(1100, 'Predicting Recession', None, None), (...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\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",
              "      <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",
              "      <th>879</th>\n",
              "      <td>16980</td>\n",
              "      <td>paul80@stthomas.edu</td>\n",
              "      <td>tayala</td>\n",
              "      <td>Nicole</td>\n",
              "      <td>Butler</td>\n",
              "      <td>4.369247</td>\n",
              "      <td>usc</td>\n",
              "      <td>founding director</td>\n",
              "      <td>bachelors degree</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>[(808, 'since January 1, 2010, NYSE', 'follows...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>880</th>\n",
              "      <td>17046</td>\n",
              "      <td>warderic@hotmail.com</td>\n",
              "      <td>hchapman</td>\n",
              "      <td>Jennifer</td>\n",
              "      <td>Bush</td>\n",
              "      <td>6.972213</td>\n",
              "      <td>NaN</td>\n",
              "      <td>head of marketing  operations for brentmark in...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>[(148, 'and are provisional and are not includ...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>881</th>\n",
              "      <td>17064</td>\n",
              "      <td>dwaynemanning@qq.com</td>\n",
              "      <td>robert66</td>\n",
              "      <td>Ronald</td>\n",
              "      <td>Anderson</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>amherst college</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>882</th>\n",
              "      <td>17100</td>\n",
              "      <td>nicholefloyd@gmail.com</td>\n",
              "      <td>burtoncheryl</td>\n",
              "      <td>David</td>\n",
              "      <td>Garrison</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>883</th>\n",
              "      <td>17118</td>\n",
              "      <td>smithmackenzie@nakanohiroyuki.com</td>\n",
              "      <td>crystal53</td>\n",
              "      <td>Nicole</td>\n",
              "      <td>Ruiz</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>NaN</td>\n",
              "      <td>consultant</td>\n",
              "      <td>NaN</td>\n",
              "      <td>california state university long beach</td>\n",
              "      <td>...</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>884 rows × 49 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-68f14015-8a28-4c19-9505-b008879a1837')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-68f14015-8a28-4c19-9505-b008879a1837 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-68f14015-8a28-4c19-9505-b008879a1837');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "       uid                              email         username   first_name  \\\n",
              "0      678      christopherknight@rmdslab.com  castrochristina  Christopher   \n",
              "1      246                  cgreene@gmail.com          gtucker     Jennifer   \n",
              "2     2232                  hthomas@gmail.com       zrodriguez      Spencer   \n",
              "3     3252                kimjanice@gmail.com        lucashall        Larry   \n",
              "4     9174           haroldguerrero@gmail.com     debbiekelley        Craig   \n",
              "..     ...                                ...              ...          ...   \n",
              "879  16980                paul80@stthomas.edu           tayala       Nicole   \n",
              "880  17046               warderic@hotmail.com         hchapman     Jennifer   \n",
              "881  17064               dwaynemanning@qq.com         robert66       Ronald   \n",
              "882  17100             nicholefloyd@gmail.com     burtoncheryl        David   \n",
              "883  17118  smithmackenzie@nakanohiroyuki.com        crystal53       Nicole   \n",
              "\n",
              "    last_name  impact_score                  company  \\\n",
              "0        Long     40.460932  university of st thomas   \n",
              "1     Robbins     29.369300  simon fraser university   \n",
              "2       Ramos     23.920091     universitas pasundan   \n",
              "3        Shah     15.879526                      NaN   \n",
              "4      Tanner     14.205337                      NaN   \n",
              "..        ...           ...                      ...   \n",
              "879    Butler      4.369247                      usc   \n",
              "880      Bush      6.972213                      NaN   \n",
              "881  Anderson      0.000000          amherst college   \n",
              "882  Garrison      0.000000                      NaN   \n",
              "883      Ruiz      0.000000                      NaN   \n",
              "\n",
              "                                             job_title              education  \\\n",
              "0                                                  NaN                    NaN   \n",
              "1                                                  ceo                    NaN   \n",
              "2                                theatre stage manager                    NaN   \n",
              "3                                              student                    NaN   \n",
              "4                                                  NaN  less than high school   \n",
              "..                                                 ...                    ...   \n",
              "879                                  founding director       bachelors degree   \n",
              "880  head of marketing  operations for brentmark in...                    NaN   \n",
              "881                                                NaN                    NaN   \n",
              "882                                                NaN                    NaN   \n",
              "883                                         consultant                    NaN   \n",
              "\n",
              "                                 university  ... filter_test  filter_spam  \\\n",
              "0                                       NaN  ...       False        False   \n",
              "1                                       NaN  ...       False        False   \n",
              "2                                       NaN  ...       False        False   \n",
              "3                          adler university  ...       False        False   \n",
              "4                                       NaN  ...       False        False   \n",
              "..                                      ...  ...         ...          ...   \n",
              "879                                     NaN  ...       False        False   \n",
              "880                                     NaN  ...       False        False   \n",
              "881                                     NaN  ...       False        False   \n",
              "882                                     NaN  ...       False        False   \n",
              "883  california state university long beach  ...       False        False   \n",
              "\n",
              "     filter_spam_2                                       project_list  \\\n",
              "0            False  [(489, 'RMDS Meetup Speakers Presentations', N...   \n",
              "1            False  [(376, 'Linear Regression Analysis of US Censu...   \n",
              "2            False  [(884, 'Bank Churn', None, None), (1251, 'Educ...   \n",
              "3            False                                                NaN   \n",
              "4            False  [(1100, 'Predicting Recession', None, None), (...   \n",
              "..             ...                                                ...   \n",
              "879          False  [(808, 'since January 1, 2010, NYSE', 'follows...   \n",
              "880          False  [(148, 'and are provisional and are not includ...   \n",
              "881          False                                                NaN   \n",
              "882          False                                                NaN   \n",
              "883          False                                                NaN   \n",
              "\n",
              "     covid_competition  news_competition  property_competition  \\\n",
              "0                False             False                 False   \n",
              "1                False             False                  True   \n",
              "2                False             False                 False   \n",
              "3                False             False                 False   \n",
              "4                False             False                 False   \n",
              "..                 ...               ...                   ...   \n",
              "879              False             False                 False   \n",
              "880              False             False                 False   \n",
              "881              False             False                 False   \n",
              "882              False             False                 False   \n",
              "883              False             False                 False   \n",
              "\n",
              "     restaurant_competition  investment_competition  recency  \n",
              "0                     False                   False        0  \n",
              "1                     False                   False        0  \n",
              "2                     False                   False        0  \n",
              "3                     False                   False        0  \n",
              "4                     False                   False        0  \n",
              "..                      ...                     ...      ...  \n",
              "879                   False                   False        0  \n",
              "880                   False                   False        0  \n",
              "881                   False                   False        0  \n",
              "882                   False                   False        0  \n",
              "883                   False                   False        0  \n",
              "\n",
              "[884 rows x 49 columns]"
            ]
          },
          "execution_count": 62,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "all_user_features = pd.read_csv('fake_all_user_features.csv')\n",
        "print(all_user_features.info())\n",
        "all_user_features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 63,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 956
        },
        "id": "Avon8RaEK7h2",
        "outputId": "74b79ec1-f893-4d63-f7fa-1cc913d8ef9e"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 829 entries, 0 to 828\n",
            "Data columns (total 11 columns):\n",
            " #   Column            Non-Null Count  Dtype  \n",
            "---  ------            --------------  -----  \n",
            " 0   uid               829 non-null    int64  \n",
            " 1   node              829 non-null    object \n",
            " 2   webinar           829 non-null    object \n",
            " 3   url               422 non-null    object \n",
            " 4   time              422 non-null    float64\n",
            " 5   time_period       422 non-null    float64\n",
            " 6   all_visits        422 non-null    object \n",
            " 7   enrolment         363 non-null    float64\n",
            " 8   all_enrol         363 non-null    object \n",
            " 9   purchased_entity  254 non-null    float64\n",
            " 10  all_orders        254 non-null    object \n",
            "dtypes: float64(4), int64(1), object(6)\n",
            "memory usage: 71.4+ KB\n",
            "None\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-ce1f0e40-1c36-4e11-bd31-f8812d8a814e\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>uid</th>\n",
              "      <th>node</th>\n",
              "      <th>webinar</th>\n",
              "      <th>url</th>\n",
              "      <th>time</th>\n",
              "      <th>time_period</th>\n",
              "      <th>all_visits</th>\n",
              "      <th>enrolment</th>\n",
              "      <th>all_enrol</th>\n",
              "      <th>purchased_entity</th>\n",
              "      <th>all_orders</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10242</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2052</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>136.0</td>\n",
              "      <td>136.0</td>\n",
              "      <td>250.0</td>\n",
              "      <td>250</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>16392</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>85.0</td>\n",
              "      <td>85.0</td>\n",
              "      <td>208.0</td>\n",
              "      <td>208</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>10248</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2058</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\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",
              "      <th>824</th>\n",
              "      <td>16374</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>http://grmds.org/user/login</td>\n",
              "      <td>1.643239e+09</td>\n",
              "      <td>1643238.0</td>\n",
              "      <td>[\"'http://grmds.org/comp/nft?utm_source=email&amp;...</td>\n",
              "      <td>111.0</td>\n",
              "      <td>111.0</td>\n",
              "      <td>265.0</td>\n",
              "      <td>265</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>825</th>\n",
              "      <td>2040</td>\n",
              "      <td>['/node/524']</td>\n",
              "      <td>['/webinar/An-Introduction-to-the-Data-Resourc...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>108.0</td>\n",
              "      <td>108.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>826</th>\n",
              "      <td>10230</td>\n",
              "      <td>['/node/376', '/node/625', '/node/375']</td>\n",
              "      <td>[]</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>272.0</td>\n",
              "      <td>272.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>827</th>\n",
              "      <td>10236</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>http://grmds.org/user/login</td>\n",
              "      <td>1.644367e+09</td>\n",
              "      <td>1644366.0</td>\n",
              "      <td>[\"'http://grmds.org/ds-volunteer-application-f...</td>\n",
              "      <td>111.0</td>\n",
              "      <td>111.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>828</th>\n",
              "      <td>2046</td>\n",
              "      <td>[]</td>\n",
              "      <td>[]</td>\n",
              "      <td>http://grmds.org/braceImpact</td>\n",
              "      <td>1.642379e+09</td>\n",
              "      <td>1642379.0</td>\n",
              "      <td>[\"'http://grmds.org/user/login'\", \"'http://grm...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>829 rows × 11 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ce1f0e40-1c36-4e11-bd31-f8812d8a814e')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ce1f0e40-1c36-4e11-bd31-f8812d8a814e button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ce1f0e40-1c36-4e11-bd31-f8812d8a814e');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "       uid                                     node  \\\n",
              "0    10242                                       []   \n",
              "1     2052                                       []   \n",
              "2    16392                                       []   \n",
              "3    10248                                       []   \n",
              "4     2058                                       []   \n",
              "..     ...                                      ...   \n",
              "824  16374                                       []   \n",
              "825   2040                            ['/node/524']   \n",
              "826  10230  ['/node/376', '/node/625', '/node/375']   \n",
              "827  10236                                       []   \n",
              "828   2046                                       []   \n",
              "\n",
              "                                               webinar  \\\n",
              "0                                                   []   \n",
              "1                                                   []   \n",
              "2                                                   []   \n",
              "3                                                   []   \n",
              "4                                                   []   \n",
              "..                                                 ...   \n",
              "824                                                 []   \n",
              "825  ['/webinar/An-Introduction-to-the-Data-Resourc...   \n",
              "826                                                 []   \n",
              "827                                                 []   \n",
              "828                                                 []   \n",
              "\n",
              "                              url          time  time_period  \\\n",
              "0                             NaN           NaN          NaN   \n",
              "1                             NaN           NaN          NaN   \n",
              "2                             NaN           NaN          NaN   \n",
              "3                             NaN           NaN          NaN   \n",
              "4                             NaN           NaN          NaN   \n",
              "..                            ...           ...          ...   \n",
              "824   http://grmds.org/user/login  1.643239e+09    1643238.0   \n",
              "825                           NaN           NaN          NaN   \n",
              "826                           NaN           NaN          NaN   \n",
              "827   http://grmds.org/user/login  1.644367e+09    1644366.0   \n",
              "828  http://grmds.org/braceImpact  1.642379e+09    1642379.0   \n",
              "\n",
              "                                            all_visits  enrolment all_enrol  \\\n",
              "0                                                  NaN        NaN       NaN   \n",
              "1                                                  NaN      136.0     136.0   \n",
              "2                                                  NaN       85.0      85.0   \n",
              "3                                                  NaN        NaN       NaN   \n",
              "4                                                  NaN        NaN       NaN   \n",
              "..                                                 ...        ...       ...   \n",
              "824  [\"'http://grmds.org/comp/nft?utm_source=email&...      111.0     111.0   \n",
              "825                                                NaN      108.0     108.0   \n",
              "826                                                NaN      272.0     272.0   \n",
              "827  [\"'http://grmds.org/ds-volunteer-application-f...      111.0     111.0   \n",
              "828  [\"'http://grmds.org/user/login'\", \"'http://grm...        NaN       NaN   \n",
              "\n",
              "     purchased_entity all_orders  \n",
              "0                 NaN        NaN  \n",
              "1               250.0        250  \n",
              "2               208.0        208  \n",
              "3                 NaN        NaN  \n",
              "4                 NaN        NaN  \n",
              "..                ...        ...  \n",
              "824             265.0        265  \n",
              "825               NaN        NaN  \n",
              "826               NaN        NaN  \n",
              "827               NaN        NaN  \n",
              "828               NaN        NaN  \n",
              "\n",
              "[829 rows x 11 columns]"
            ]
          },
          "execution_count": 63,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "match_df = pd.read_csv('match_df.csv')\n",
        "print(match_df.info())\n",
        "match_df"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AWLVBMoYCK9M"
      },
      "source": [
        "# **User Sim**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 55,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tz8j-R8_ssZU",
        "outputId": "07a5c08b-4a6e-4a2a-e3d1-3381bf69fa0f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: pyjarowinkler in /usr/local/lib/python3.7/dist-packages (1.8)\n"
          ]
        }
      ],
      "source": [
        "pip install pyjarowinkler"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 56,
      "metadata": {
        "id": "FS7a_09wCK9M"
      },
      "outputs": [],
      "source": [
        "from pyjarowinkler import distance   \n",
        "import re\n",
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "import itertools\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import ast\n",
        "from ast import literal_eval\n",
        "from itertools import combinations\n",
        "import pickle\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 57,
      "metadata": {
        "id": "edBV6yQPUl-B"
      },
      "outputs": [],
      "source": [
        "# Methods for generate similarities\n",
        "class recom_user(object):\n",
        "    def __init__(self):\n",
        "        self.interest_list = ['Fraud_Detection', 'Risk_Scoring', 'Healthcare', 'Internet_Search', 'Marketing_Effectiveness',\n",
        "                              'Website_Recommendations', 'Image_Recognition', 'Speech_Recognition',\n",
        "                              'Airline_Route_Planning', 'Price_Analytics', 'Supply_Chain_Optimization',\n",
        "                              'Talent_Acquisition_Analytics', 'Environment_Analytics', 'Epidemiology',\n",
        "                              'Social_Policy', 'Evaluation_and_Assessment']\n",
        "        self.vectorizer = TfidfVectorizer()\n",
        "        \n",
        "    def compare(self, info1, info2):\n",
        "        results = {}\n",
        "        \n",
        "        #email extension sim\n",
        "        results['email_sim'] = self.email_extension(info1['email'], info2['email'])\n",
        "        \n",
        "        #University company job_title education major sim\n",
        "        user_info_list = ['university','company','job_title','education','major']\n",
        "        for l in user_info_list:\n",
        "            results[l] = self.simple_text(info1[l], info2[l])\n",
        "        \n",
        "        # interest sim\n",
        "        results['interest_sim'] = self.cos_sim([info1[x] if info1[x] != None else 0 for x in self.interest_list], \n",
        "                                               [info2[x] if info2[x] != None else 0 for x in self.interest_list])\n",
        "        # followers\n",
        "        results['follower_sim'] = info2['num_follower']\n",
        "\n",
        "        # impact score\n",
        "        results['impact_score_sim'] = info2['impact_score']        \n",
        "\n",
        "        # recency\n",
        "        results['recency_sim'] = info2['recency']\n",
        "\n",
        "        # news competition\n",
        "        competition_list = ['covid_competition','news_competition','property_competition','restaurant_competition',\n",
        "                            'investment_competition']\n",
        "        for l in competition_list:\n",
        "            results[l] = info1[l] & info2[l]\n",
        "\n",
        "        # project part\n",
        "        if (info1['project_list']==None or info2['project_list']==None or \n",
        "            len(info1['project_list']) ==0 or len(info2['project_list'])==0):\n",
        "            results['proj_title_sim'],results['proj_des_sim'],results['proj_language_sim'] = None, None, None\n",
        "            return results\n",
        "     \n",
        "        # project title sim\n",
        "        title_list1 = [x[1] for x in info1['project_list']]\n",
        "        title_list2 = [x[1] for x in info2['project_list']]\n",
        "        pairs = list(itertools.product(title_list1, title_list2))\n",
        "        results['proj_title_sim'] = sum(map(lambda x : self.simple_text(x[0],x[1]), pairs)) / len(pairs)\n",
        "        \n",
        "        # project language sim\n",
        "        lan_list1 = [x[3] for x in info1['project_list'] if x[3] != None]\n",
        "        lan_list2 = [x[3] for x in info2['project_list'] if x[3] != None]\n",
        "        if len(lan_list1) == 0 or len(lan_list2) == 0:\n",
        "            results['proj_language_sim'] = None\n",
        "        else:\n",
        "            pairs = list(itertools.product(lan_list1, lan_list2))\n",
        "            results['proj_language_sim'] = sum(map(lambda x : x[0].lower()==x[1].lower(), pairs)) / len(pairs) * 100\n",
        "            \n",
        "        # project description sim\n",
        "        des_list1 = [x[2] for x in info1['project_list'] if x[2] != None]\n",
        "        des_list2 = [x[2] for x in info2['project_list'] if x[2] != None]\n",
        "        if len(des_list1) == 0 or len(des_list2) == 0:\n",
        "            results['proj_des_sim'] = None\n",
        "        else:\n",
        "            results['proj_des_sim'] = self.long_text(des_list1, des_list2)\n",
        "        \n",
        "        return results\n",
        "        \n",
        "    def email_extension(self,u1,u2):\n",
        "        # Judge if these two email address have the same suffix\n",
        "        if u1 in [None,''] or u2 in [None,'']:\n",
        "            return None\n",
        "        try:\n",
        "            return (u1.split('@')[1].split('.')[1] == u2.split('@')[1].split('.')[1]) * 100\n",
        "        except:\n",
        "            return None\n",
        "        \n",
        "    def simple_text(self,u1,u2):\n",
        "        if u1 in [None,''] or u2 in [None,'']:\n",
        "            return None\n",
        "        #u1 = re.sub(r'[^\\w]', '', u1).lower()\n",
        "        #u2 = re.sub(r'[^\\w]', '', u2).lower()\n",
        "        return distance.get_jaro_distance(u1, u2) * 100\n",
        "    \n",
        "    def cos_sim(self,u1,u2):\n",
        "    #   return cosine_similarity([u1], [u2])[0][0] * 100\n",
        "        u1 = np.array(u1)\n",
        "        u2 = np.array(u2)\n",
        "        # numpy's cosine similarity\n",
        "        cosine_sim = np.dot(u1,u2) / (np.sqrt(np.dot(u1,u1)) * np.sqrt(np.dot(u2,u2)))\n",
        "        if np.isnan(cosine_sim):\n",
        "          return None\n",
        "        else:\n",
        "          return cosine_sim * 100\n",
        "    \n",
        "    def long_text(self,doc1,doc2):\n",
        "        # Vectorize long text information and calculate their cosine similarity score\n",
        "        #vectorizer = TfidfVectorizer()\n",
        "        vectors = self.vectorizer.fit_transform(doc1+doc2)\n",
        "        dense = vectors.todense()\n",
        "        dense1 = dense[:len(doc1)]\n",
        "        dense2 = dense[len(doc1):]\n",
        "        pairs = list(itertools.product(dense1, dense2))\n",
        "        return sum(map(lambda x : cosine_similarity(x[0], x[1])[0][0] , pairs)) / len(pairs) * 100\n",
        "    \n",
        "        "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 103,
      "metadata": {
        "id": "XDOb_xBgUCdS"
      },
      "outputs": [],
      "source": [
        "# Generate dependent varibles, as matched/unmatched\n",
        "# Matching or not is decided by wheather or not the users have some common activities:\n",
        "# Visiting a same project/webinar/course/other pages, Taking a same order...\n",
        "\n",
        "class user_sim_evaluator(object):\n",
        "    def __init__(self,n_node, n_webinar,match_file,feature_file):\n",
        "        match_df = pd.read_csv(match_file,index_col=0)\n",
        "        all_user_features = pd.read_csv(feature_file,index_col=0)\n",
        "        all_user_features = self.clean_features(all_user_features)\n",
        "        match_df = self.clean_match(match_df)\n",
        "        self.recom = recom_user()\n",
        "        self.pair_wise, self.common_ids = self.evaluate_all(n_node, n_webinar,all_user_features,match_df) \n",
        "        \n",
        "    def evaluate_all(self, n_node, n_webinar, all_user_features,match_df):\n",
        "        # Filter out users with at least one visit in project/webinar\n",
        "        valid_list = match_df[(match_df.node.str.find('/node/')>-1) | (match_df.webinar.str.find('/webinar/')>-1)]\n",
        "        # Turn the string of node and webinar into list syntax\n",
        "        valid_list[['node', 'webinar']] = valid_list[['node', 'webinar']].applymap(literal_eval)\n",
        "        valid_ids = valid_list.index.to_list()\n",
        "        common_ids = list(set(valid_ids).intersection(all_user_features.index))\n",
        "        # In case ids are mixed\n",
        "\n",
        "        # formatting features for evaluation \n",
        "        filter = all_user_features.index.isin(common_ids)\n",
        "        all_features = all_user_features[filter]\n",
        "        all_features = all_features.where(pd.notnull(all_features), None)\n",
        "        valid_list = valid_list.where(pd.notnull(valid_list), None)\n",
        "              \n",
        "\n",
        "        pair_wise = []\n",
        "        count = 0\n",
        "        for tuples in combinations(common_ids, 2):\n",
        "            count += 1\n",
        "            r1 = valid_list.loc[tuples[0],:]\n",
        "            r2 = valid_list.loc[tuples[1],:]\n",
        "            match = []\n",
        "            match.append(not set(r1['node'][:n_node]+r1['webinar'][:n_webinar]).isdisjoint(r2['node'][:n_node]+r2['webinar'][:n_webinar]))\n",
        "\n",
        "            # Checking common in courses, orders, visits\n",
        "            activities_list = ['all_enrol','all_orders','all_visits']\n",
        "            for l in activities_list:\n",
        "              activity_1 = valid_list.loc[tuples[0],l]\n",
        "              activity_2 = valid_list.loc[tuples[1],l]\n",
        "              if activity_1 == None or activity_2 ==None:\n",
        "                match.append(False)\n",
        "              elif len(set(activity_1)&set(activity_2))>0:\n",
        "                match.append(True)\n",
        "              else:\n",
        "                match.append(False)\n",
        "\n",
        "            \n",
        "            # If there is one match, users have common activity\n",
        "            match = sum(match)\n",
        "            if match >1: \n",
        "                match = 1 \n",
        "            uinfo_1 = all_features.loc[tuples[0],:]\n",
        "            uinfo_2 = all_features.loc[tuples[1],:]\n",
        "            sim_score = self.recom.compare(uinfo_1,uinfo_2)                                                                                            \n",
        "            pair_wise.append({'user1': tuples[0], 'user2': tuples[1], 'match': match, 'sim_score': sim_score})\n",
        "\n",
        "        # Create a pickle file to save the data\n",
        "        with open(\"pairwise_compare_dict.pickle\", \"wb\",) as myFile:\n",
        "            pickle.dump(pair_wise, myFile)\n",
        "    \n",
        "        return pair_wise, common_ids\n",
        "\n",
        "    def literal_return(self,val):\n",
        "        # Turn the string into list syntax\n",
        "        try:\n",
        "            return ast.literal_eval(val)\n",
        "        except (ValueError, SyntaxError) as e:\n",
        "            return None\n",
        "\n",
        "    def str_to_list(self,string):\n",
        "        # Turn the string into list syntax\n",
        "        if str(string) not in ['NaN','nan','None']:\n",
        "            return string.strip('][').split(', ')\n",
        "        else:\n",
        "            return None\n",
        "\n",
        "    def clean_features(self,all_user_features):\n",
        "        # Clean the text data\n",
        "        all_user_features[['project_list']] = all_user_features[['project_list']].applymap(self.literal_return)\n",
        "        clean_list = ['company','job_title','education','major','university']\n",
        "        for l in clean_list:\n",
        "            all_user_features[l] = all_user_features[l].apply(lambda x: re.sub(r'[^\\w\\s]', '', str(x)).lower())\n",
        "        return all_user_features\n",
        "\n",
        "    def clean_match(self,match_df):\n",
        "        clean_list = ['all_enrol','all_visits','all_orders']\n",
        "        for l in clean_list:\n",
        "            match_df[[l]] = match_df[[l]].applymap(self.str_to_list)\n",
        "        return match_df\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 104,
      "metadata": {
        "id": "6R-hRYvSVbBs"
      },
      "outputs": [],
      "source": [
        "user_sim = user_sim_evaluator(5,3,'match_df.csv','fake_all_user_features.csv') #Numbers up to change"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 105,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pk-u8GorX_q2",
        "outputId": "82f99cb1-593a-437b-cfe2-ab2081e7d71a"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "30381"
            ]
          },
          "execution_count": 105,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "pair_wise = user_sim.pair_wise\n",
        "len(pair_wise)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 73,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sEiGLMXZYICv",
        "outputId": "fde672db-2565-4b2d-c044-2f08de20cd7f"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "0.32250419670188607"
            ]
          },
          "execution_count": 73,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "i = 0\n",
        "for n in range(len(pair_wise)):\n",
        "   i += pair_wise[n]['match']\n",
        "\n",
        "i/len(pair_wise)\n",
        "# See how much matches we could get"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xvBLMv_c0SnT"
      },
      "source": [
        "# **Model retraining and update**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 89,
      "metadata": {
        "id": "oDuxmGhUyVBE"
      },
      "outputs": [],
      "source": [
        "# -*- coding: utf-8 -*-\n",
        "\"\"\"\n",
        "Created on Thu Jul 23 11:40:08 2020\n",
        "@author: jidong\n",
        "\"\"\"\n",
        "import pandas as pd\n",
        "from sklearn import model_selection\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "class model_retrain(object):\n",
        "    def __init__(self,final_dict,all_user_features):\n",
        "        model, res = self.model_retrain_test(final_dict)\n",
        "        model.to_csv('model_weights.csv')\n",
        "        res.to_csv('model_accuracy.csv')\n",
        "\n",
        "        self.recom = recom_user()\n",
        "        all_user_features = self.clean_features(all_user_features)\n",
        "        self.user_sim_table = self.update_user_recom_table(all_user_features, model)\n",
        "        self.user_sim_table.to_csv('user_sim_table.csv')\n",
        "\n",
        "    def model_retrain_test(self,data):\n",
        "        for elem in data:\n",
        "            score = elem['sim_score']\n",
        "            key_list = score.keys()\n",
        "            for key in key_list:\n",
        "                elem[key] = score[key]\n",
        "            elem.pop('sim_score')\n",
        "        \n",
        "        pair_df = pd.DataFrame(data)\n",
        "        col_name = list(pair_df.columns)[3:]\n",
        "        \n",
        "        # Clean the pairwise data\n",
        "        process_pair_df = pair_df.dropna(thresh = 6)\n",
        "        process_pair_df = process_pair_df.fillna(0)\n",
        "        process_pair_df['match'] = pair_df['match'].astype(int)\n",
        "        x = process_pair_df.iloc[:, 3:].to_numpy()\n",
        "        y = np.array(process_pair_df['match'])\n",
        "        \n",
        "        # Utilize LR model to get the weight of each similarity features\n",
        "        # Dependent variables are users' match\n",
        "        # Independent variables are users' similarity scores\n",
        "        model = LogisticRegression(solver='liblinear', multi_class='ovr')\n",
        "        kfold = model_selection.KFold(n_splits=10)\n",
        "        cv = model_selection.cross_validate(model, x, y, cv=kfold, scoring='accuracy', return_estimator=True)\n",
        "        cv_results = cv['test_score']\n",
        "        best_model = cv['estimator'][np.argmax(cv_results)]\n",
        "        \n",
        "        # Generate results df\n",
        "        results_table = pd.DataFrame()\n",
        "        results_table['accuracy'] = cv_results\n",
        "        results_table['timestamp'] = pd.Timestamp.today()\n",
        "        best_model.fit(x,y)\n",
        "        model_para = np.absolute(best_model.coef_[0])\n",
        "        model_para = model_para / model_para.sum()\n",
        "        model_para = pd.DataFrame(zip(col_name, model_para), columns = ['features','weights'])\n",
        "        model_para['timestamp'] = pd.Timestamp.today()\n",
        "            \n",
        "        return model_para, results_table\n",
        "\n",
        "        # Periodically update the recommendation score table\n",
        "        # Calculate total similarity score of each user by using the weights in LR model\n",
        "    def update_user_recom_table(self,data, model):\n",
        "        results = []\n",
        "        count = 0\n",
        "        user_list = data.index.to_list()\n",
        "        weight = np.array(model['weights'])\n",
        "        for user1 in user_list:\n",
        "            sim_user = []\n",
        "            for user2 in user_list:\n",
        "                if user1 != user2:\n",
        "                    count += 1\n",
        "                    if count%100000 == 0:\n",
        "                        print(count)\n",
        "                    compare = self.recom.compare(data.loc[user1,:], data.loc[user2,:])\n",
        "                    compare = {key:0 if str(value) in ['nan','None','NaN'] else value for key,value in compare.items()}\n",
        "                    score = np.dot(np.array(list(compare.values())),weight)\n",
        "                    sim_user.append({'user': user1, 'sim_user': user2, 'sim_score':score})\n",
        "            sim_user = sorted(sim_user, key = lambda x: x['sim_score'], reverse = True)[:10] # Could be 50\n",
        "            results.extend(sim_user)\n",
        "        results = pd.DataFrame(results)\n",
        "        return results\n",
        "\n",
        "    def clean_features(self,all_user_features):\n",
        "        # Clean the text data\n",
        "        all_user_features[['project_list']] = all_user_features[['project_list']].applymap(self.literal_return)\n",
        "        clean_list = ['company','job_title','education','major','university']\n",
        "        for l in clean_list:\n",
        "            all_user_features[l] = all_user_features[l].apply(lambda x: re.sub(r'[^\\w\\s]', '', str(x)).lower())\n",
        "        return all_user_features\n",
        "    \n",
        "    def literal_return(self,val):\n",
        "        # Turn the string into list syntax\n",
        "        try:\n",
        "            return ast.literal_eval(val)\n",
        "        except (ValueError, SyntaxError) as e:\n",
        "            return None"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 93,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dj5AbenaybE8",
        "outputId": "1d82ce48-a9bb-41ce-dcb8-e4194578d84f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "100000\n",
            "200000\n",
            "300000\n",
            "400000\n",
            "500000\n",
            "600000\n",
            "700000\n"
          ]
        }
      ],
      "source": [
        "final_dict = pd.read_pickle('pairwise_compare_dict.pickle')\n",
        "all_user_features = pd.read_csv('fake_all_user_features.csv',index_col=0)\n",
        "retrain = model_retrain(final_dict, all_user_features)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 94,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "kHJ2-FNiFVHg",
        "outputId": "3b0b58a1-b1ab-424d-f3ab-4ebef28b0a45"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-d6328626-2aa6-4ed8-a7ca-c6bfec711f2d\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Unnamed: 0</th>\n",
              "      <th>user</th>\n",
              "      <th>sim_user</th>\n",
              "      <th>sim_score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>678</td>\n",
              "      <td>6552</td>\n",
              "      <td>1.517091</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>678</td>\n",
              "      <td>3396</td>\n",
              "      <td>1.451002</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>678</td>\n",
              "      <td>16890</td>\n",
              "      <td>1.433840</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>678</td>\n",
              "      <td>9990</td>\n",
              "      <td>1.423084</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>678</td>\n",
              "      <td>8964</td>\n",
              "      <td>1.217707</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8835</th>\n",
              "      <td>8835</td>\n",
              "      <td>17118</td>\n",
              "      <td>9174</td>\n",
              "      <td>0.511951</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8836</th>\n",
              "      <td>8836</td>\n",
              "      <td>17118</td>\n",
              "      <td>9120</td>\n",
              "      <td>0.507265</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8837</th>\n",
              "      <td>8837</td>\n",
              "      <td>17118</td>\n",
              "      <td>16230</td>\n",
              "      <td>0.497397</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8838</th>\n",
              "      <td>8838</td>\n",
              "      <td>17118</td>\n",
              "      <td>492</td>\n",
              "      <td>0.492900</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8839</th>\n",
              "      <td>8839</td>\n",
              "      <td>17118</td>\n",
              "      <td>10008</td>\n",
              "      <td>0.488623</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>8840 rows × 4 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d6328626-2aa6-4ed8-a7ca-c6bfec711f2d')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-d6328626-2aa6-4ed8-a7ca-c6bfec711f2d button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-d6328626-2aa6-4ed8-a7ca-c6bfec711f2d');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "      Unnamed: 0   user  sim_user  sim_score\n",
              "0              0    678      6552   1.517091\n",
              "1              1    678      3396   1.451002\n",
              "2              2    678     16890   1.433840\n",
              "3              3    678      9990   1.423084\n",
              "4              4    678      8964   1.217707\n",
              "...          ...    ...       ...        ...\n",
              "8835        8835  17118      9174   0.511951\n",
              "8836        8836  17118      9120   0.507265\n",
              "8837        8837  17118     16230   0.497397\n",
              "8838        8838  17118       492   0.492900\n",
              "8839        8839  17118     10008   0.488623\n",
              "\n",
              "[8840 rows x 4 columns]"
            ]
          },
          "execution_count": 94,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "user_sim_table = retrain.user_sim_table\n",
        "user_sim_table"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oMrQQpLwVogO"
      },
      "source": [
        "# **Evaluation**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 123,
      "metadata": {
        "id": "RNBRjJvcDOpn"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "from sklearn import model_selection\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "import sklearn\n",
        "import sklearn.metrics as metrics\n",
        "import sklearn.metrics as metrics\n",
        "from sklearn.metrics import accuracy_score\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "class evaluation(object):\n",
        "  def __init__(self, filename):\n",
        "    final_dict = pd.read_pickle(filename)\n",
        "    X, y, model = self.cross_validate(final_dict)\n",
        "    self.y_pred_prob_test, self.y_pred = self.model_evaluate(X,y,model)\n",
        "\n",
        "\n",
        "  def cross_validate(self,data):\n",
        "    for elem in data:\n",
        "      score = elem['sim_score']\n",
        "      key_list = score.keys()\n",
        "      for key in key_list:\n",
        "          elem[key] = score[key]\n",
        "      elem.pop('sim_score')\n",
        "\n",
        "    pair_df = pd.DataFrame(data)\n",
        "    col_name = list(pair_df.columns)[3:]\n",
        "\n",
        "    process_pair_df = pair_df.dropna(thresh = 6)\n",
        "    process_pair_df = process_pair_df.fillna(0)\n",
        "    process_pair_df['match'] = pair_df['match'].astype(int)\n",
        "    X = process_pair_df.iloc[:, 3:].to_numpy()\n",
        "    y = np.array(process_pair_df['match'])\n",
        "\n",
        "    model = LogisticRegression(solver='liblinear', multi_class='ovr')\n",
        "    kfold = model_selection.KFold(n_splits=10)\n",
        "    cv = model_selection.cross_validate(model, X, y, cv=kfold, scoring='accuracy', return_estimator=True)\n",
        "    cv_results = cv['test_score']\n",
        "    best_model = cv['estimator'][np.argmax(cv_results)]\n",
        "    print('Cross validation results',cv_results.mean())\n",
        "\n",
        "    return X,y,model\n",
        "\n",
        "  def model_evaluate(self,X,y,model):\n",
        "    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)\n",
        "    model.fit(X_train,y_train)\n",
        "    y_pred_prob_test = model.predict_proba(X_test)\n",
        "    y_pred = model.predict(X_test)\n",
        "\n",
        "    #Confusion Matrix\n",
        "    confusion_matrix_census=sklearn.metrics.confusion_matrix(y_test, y_pred, labels=None, sample_weight=None, normalize=None)\n",
        "    tn, fp, fn, tp = confusion_matrix_census.ravel()\n",
        "    print(confusion_matrix_census)\n",
        "\n",
        "    #Precision and Recall\n",
        "    precision=tp/(tp+fp)\n",
        "    recall=tp/(tp+fn)\n",
        "    print('Precision is ',precision,' , Recall is ',recall)\n",
        "\n",
        "    #f1 score\n",
        "    f1_score_cen=sklearn.metrics.f1_score(y_test, y_pred,average='binary')\n",
        "    f1_score_cal=2*precision*recall/(precision+recall)\n",
        "    print('built-in f1 score is ',f1_score_cen)\n",
        "    print('calculated f1 score is ',f1_score_cal)\n",
        "\n",
        "\n",
        "\n",
        "    #Accuracy score and AUC score\n",
        "    accuracy=accuracy_score(y_test, y_pred)\n",
        "    print('Accuracy score is ',accuracy)\n",
        "\n",
        "    y_pred_prob = model.predict_proba(X_test)[:,1]\n",
        "    fpr, tpr, threshold = metrics.roc_curve(y_test, y_pred_prob)\n",
        "    roc_auc = metrics.auc(fpr, tpr)\n",
        "    print('AUC score is ',roc_auc)\n",
        "\n",
        "\n",
        "    #ROC\n",
        "    # method I: plt\n",
        "    plt.title('Receiver Operating Characteristic')\n",
        "    plt.plot(fpr, tpr, 'b', label = 'AUC = %0.2f' % roc_auc)\n",
        "    plt.legend(loc = 'lower right')\n",
        "    plt.plot([0, 1], [0, 1],'r--')\n",
        "    plt.xlim([0, 1])\n",
        "    plt.ylim([0, 1])\n",
        "    plt.ylabel('True Positive Rate')\n",
        "    plt.xlabel('False Positive Rate')\n",
        "    plt.show()\n",
        "\n",
        "    return y_pred_prob_test, y_pred\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 124,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 434
        },
        "id": "S4OFX3N7FGuq",
        "outputId": "d4cd84f5-1434-416e-d1ab-8506bff11291"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Cross validation results 0.6772618240685441\n",
            "[[6782   28]\n",
            " [3195   21]]\n",
            "Precision is  0.42857142857142855  , Recall is  0.0065298507462686565\n",
            "built-in f1 score is  0.012863705972434915\n",
            "calculated f1 score is  0.012863705972434915\n",
            "Accuracy score is  0.6785358069020546\n",
            "AUC score is  0.5570836392559961\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "evaluate = evaluation('pairwise_compare_dict.pickle')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 125,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xPa-KCZgGzuX",
        "outputId": "945c758a-a39e-4686-c013-0a839009d510"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "0    9419\n",
              "1     607\n",
              "dtype: int64"
            ]
          },
          "execution_count": 125,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y_pred_prob = evaluate.y_pred_prob_test\n",
        "y_prob = []\n",
        "for num in range(len(y_pred_prob)):\n",
        "  y_prob.append(y_pred_prob[num][1])\n",
        "\n",
        "results = pd.Series(y_prob) \n",
        "y_pred = results > 0.40  # Change threshold here\n",
        "y_pred = y_pred*1\n",
        "\n",
        "y_pred.value_counts()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "w9VY68QbInvn"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "colab": {
      "collapsed_sections": [
        "5dugV2MudYDh",
        "lLmycDejCK9J"
      ],
      "name": "user_sim_open_source_complete.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.8.8"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}