mishima_syk13_handson1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "mishima.syk13_hanson1.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/iwatobipen/playground/blob/master/mishima_syk13_hanson1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "Lp4mIUR58Zpf",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "##Google colabでの作業の場合\n",
        "PlayGroundモードで開きます。\n",        
        "##GPUを使いたい場合\n",
        "- ランタイム＝＞ランタイムのタイプを変更でGPUにしてください。\n",
        "\n",
        "## 最初に環境を作ります。\n",
        "- ColabのPythonのバージョンが3.6.7だったのでMinicondaも3.6を入れます。\n",
        "- 次いでconda commandでrdkit  をインストールします。\n",
        "　＝＞そこそこ時間かかります。\n",
        "- pytorchなどGeneratorに必要なパッケージ群を入れます\n",
        "- システムPathにCondaのPathを入れることでRDKITを呼び出せるようにします。\n",
        "- 手元の環境で全部終わるのに１０分以上かかる感じなのでこれを使う場合は早めに実行しておきましょう。"
      ]
    },
    {
      "metadata": {
        "id": "iuxmNBQQ5c_F",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 3842
        },
        "outputId": "53dde51f-6536-4068-e340-b219da21d499"
      },
      "cell_type": "code",
      "source": [
        "# install conda\n",
        "!wget https://repo.continuum.io/miniconda/Miniconda3-4.4.10-Linux-x86_64.sh\n",
        "!chmod ./Miniconda3-4.4.10-Linux-x86_64.sh\n",
        "!time bash ./Miniconda3-4.4.10-Linux-x86_64.sh -b -f -p /usr/local\n",
        "# install rdkit via conda command\n",
        "!time conda install -c conda-forge rdkit -y\n",
        "# install pytorch via conda command\n",
        "!time conda install -c pytorch pytorch==0.3.0 -y\n",
        "# install sklearn via conda command\n",
        "!time conda install -c conda-forge scikit-learn -y\n",
        "!time conda install -c conda-forge pexpect  -y"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2019-01-03 04:58:55--  https://repo.continuum.io/miniconda/Miniconda3-4.4.10-Linux-x86_64.sh\n",
            "Resolving repo.continuum.io (repo.continuum.io)... 104.16.19.10, 104.16.18.10, 2606:4700::6810:130a, ...\n",
            "Connecting to repo.continuum.io (repo.continuum.io)|104.16.19.10|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 58304693 (56M) [application/x-sh]\n",
            "Saving to: ‘Miniconda3-4.4.10-Linux-x86_64.sh.1’\n",
            "\n",
            "Miniconda3-4.4.10-L 100%[===================>]  55.60M   107MB/s    in 0.5s    \n",
            "\n",
            "2019-01-03 04:58:55 (107 MB/s) - ‘Miniconda3-4.4.10-Linux-x86_64.sh.1’ saved [58304693/58304693]\n",
            "\n",
            "chmod: missing operand after ‘./Miniconda3-4.4.10-Linux-x86_64.sh’\n",
            "Try 'chmod --help' for more information.\n",
            "PREFIX=/usr/local\n",
            "installing: python-3.6.4-hc3d631a_1 ...\n",
            "Python 3.6.4 :: Anaconda, Inc.\n",
            "installing: ca-certificates-2017.08.26-h1d4fec5_0 ...\n",
            "installing: conda-env-2.6.0-h36134e3_1 ...\n",
            "installing: libgcc-ng-7.2.0-h7cc24e2_2 ...\n",
            "installing: libstdcxx-ng-7.2.0-h7a57d05_2 ...\n",
            "installing: libffi-3.2.1-hd88cf55_4 ...\n",
            "installing: ncurses-6.0-h9df7e31_2 ...\n",
            "installing: openssl-1.0.2n-hb7f436b_0 ...\n",
            "installing: tk-8.6.7-hc745277_3 ...\n",
            "installing: xz-5.2.3-h55aa19d_2 ...\n",
            "installing: yaml-0.1.7-had09818_2 ...\n",
            "installing: zlib-1.2.11-ha838bed_2 ...\n",
            "installing: libedit-3.1-heed3624_0 ...\n",
            "installing: readline-7.0-ha6073c6_4 ...\n",
            "installing: sqlite-3.22.0-h1bed415_0 ...\n",
            "installing: asn1crypto-0.24.0-py36_0 ...\n",
            "installing: certifi-2018.1.18-py36_0 ...\n",
            "installing: chardet-3.0.4-py36h0f667ec_1 ...\n",
            "installing: idna-2.6-py36h82fb2a8_1 ...\n",
            "installing: pycosat-0.6.3-py36h0a5515d_0 ...\n",
            "installing: pycparser-2.18-py36hf9f622e_1 ...\n",
            "installing: pysocks-1.6.7-py36hd97a5b1_1 ...\n",
            "installing: ruamel_yaml-0.15.35-py36h14c3975_1 ...\n",
            "installing: six-1.11.0-py36h372c433_1 ...\n",
            "installing: cffi-1.11.4-py36h9745a5d_0 ...\n",
            "installing: setuptools-38.4.0-py36_0 ...\n",
            "installing: cryptography-2.1.4-py36hd09be54_0 ...\n",
            "installing: wheel-0.30.0-py36hfd4bba0_1 ...\n",
            "installing: pip-9.0.1-py36h6c6f9ce_4 ...\n",
            "installing: pyopenssl-17.5.0-py36h20ba746_0 ...\n",
            "installing: urllib3-1.22-py36hbe7ace6_0 ...\n",
            "installing: requests-2.18.4-py36he2e5f8d_1 ...\n",
            "installing: conda-4.4.10-py36_0 ...\n",
            "unlinking: ca-certificates-2018.11.29-ha4d7672_0\n",
            "unlinking: certifi-2018.11.29-py36_1000\n",
            "unlinking: openssl-1.0.2p-h470a237_1\n",
            "unlinking: xz-5.2.4-h470a237_1\n",
            "installation finished.\n",
            "WARNING:\n",
            "    You currently have a PYTHONPATH environment variable set. This may cause\n",
            "    unexpected behavior when running the Python interpreter in Miniconda3.\n",
            "    For best results, please verify that your PYTHONPATH only points to\n",
            "    directories of packages that are compatible with the Python interpreter\n",
            "    in Miniconda3: /usr/local\n",
            "\n",
            "real\t0m18.420s\n",
            "user\t0m16.183s\n",
            "sys\t0m3.988s\n",
            "Solving environment: - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\bdone\n",
            "\n",
            "\n",
            "==> WARNING: A newer version of conda exists. <==\n",
            "  current version: 4.4.10\n",
            "  latest version: 4.5.12\n",
            "\n",
            "Please update conda by running\n",
            "\n",
            "    $ conda update -n base conda\n",
            "\n",
            "\n",
            "\n",
            "## Package Plan ##\n",
            "\n",
            "  environment location: /usr/local\n",
            "\n",
            "  added / updated specs: \n",
            "    - rdkit\n",
            "\n",
            "\n",
            "The following packages will be UPDATED:\n",
            "\n",
            "    ca-certificates: 2017.08.26-h1d4fec5_0 --> 2018.11.29-ha4d7672_0 conda-forge\n",
            "    certifi:         2018.1.18-py36_0      --> 2018.11.29-py36_1000  conda-forge\n",
            "    openssl:         1.0.2n-hb7f436b_0     --> 1.0.2p-h470a237_1     conda-forge\n",
            "    xz:              5.2.3-h55aa19d_2      --> 5.2.4-h470a237_1      conda-forge\n",
            "\n",
            "Preparing transaction: | \b\bdone\n",
            "Verifying transaction: - \b\b\\ \b\bdone\n",
            "Executing transaction: / \b\b- \b\bdone\n",
            "\n",
            "real\t0m34.177s\n",
            "user\t0m32.817s\n",
            "sys\t0m0.926s\n",
            "Solving environment: - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\bdone\n",
            "\n",
            "\n",
            "==> WARNING: A newer version of conda exists. <==\n",
            "  current version: 4.4.10\n",
            "  latest version: 4.5.12\n",
            "\n",
            "Please update conda by running\n",
            "\n",
            "    $ conda update -n base conda\n",
            "\n",
            "\n",
            "\n",
            "## Package Plan ##\n",
            "\n",
            "  environment location: /usr/local\n",
            "\n",
            "  added / updated specs: \n",
            "    - pytorch==0.4.0\n",
            "\n",
            "\n",
            "The following packages will be downloaded:\n",
            "\n",
            "    package                    |            build\n",
            "    ---------------------------|-----------------\n",
            "    cudatoolkit-8.0            |                3       322.4 MB\n",
            "    pytorch-0.4.0              |py36_cuda8.0.61_cudnn7.1.2_1       390.8 MB  pytorch\n",
            "    ------------------------------------------------------------\n",
            "                                           Total:       713.2 MB\n",
            "\n",
            "The following NEW packages will be INSTALLED:\n",
            "\n",
            "    cudatoolkit: 8.0-3                                        \n",
            "\n",
            "The following packages will be DOWNGRADED:\n",
            "\n",
            "    pytorch:     0.4.1-py36_py35_py27__9.0.176_7.1.2_2 pytorch --> 0.4.0-py36_cuda8.0.61_cudnn7.1.2_1 pytorch\n",
            "\n",
            "\n",
            "Downloading and Extracting Packages\n",
            "cudatoolkit 8.0: 100% 1.0/1 [02:13<00:00, 704.22s/it]               \n",
            "pytorch 0.4.0: 100% 1.0/1 [03:07<00:00, 187.04s/it]               \n",
            "Preparing transaction: | \b\bdone\n",
            "Verifying transaction: - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\bdone\n",
            "Executing transaction: \\ \b\b| \b\b/ \b\bdone\n",
            "\n",
            "real\t5m39.892s\n",
            "user\t5m30.129s\n",
            "sys\t0m6.056s\n",
            "Solving environment: - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\bdone\n",
            "\n",
            "\n",
            "==> WARNING: A newer version of conda exists. <==\n",
            "  current version: 4.4.10\n",
            "  latest version: 4.5.12\n",
            "\n",
            "Please update conda by running\n",
            "\n",
            "    $ conda update -n base conda\n",
            "\n",
            "\n",
            "\n",
            "## Package Plan ##\n",
            "\n",
            "  environment location: /usr/local\n",
            "\n",
            "  added / updated specs: \n",
            "    - scikit-learn\n",
            "\n",
            "\n",
            "The following packages will be UPDATED:\n",
            "\n",
            "    blas:         1.0-mkl               --> 1.1-openblas                          conda-forge\n",
            "    numpy:        1.15.0-py36h1b885b7_0 --> 1.15.4-py36_blas_openblashb06ca3d_0   conda-forge [blas_openblas]\n",
            "    scikit-learn: 0.19.1-py36hedc7406_0 --> 0.20.2-py36_blas_openblash00c3548_400 conda-forge [blas_openblas]\n",
            "    scipy:        1.1.0-py36hc49cb51_0  --> 1.2.0-py36_blas_openblashb06ca3d_200  conda-forge [blas_openblas]\n",
            "\n",
            "The following packages will be DOWNGRADED:\n",
            "\n",
            "    numpy-base:   1.15.0-py36h3dfced4_0 --> 1.14.3-py36h0ea5e3f_1                            \n",
            "\n",
            "Preparing transaction: / \b\b- \b\b\\ \b\bdone\n",
            "Verifying transaction: / \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\bdone\n",
            "Executing transaction: | \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\bdone\n",
            "\n",
            "real\t0m47.676s\n",
            "user\t0m43.709s\n",
            "sys\t0m2.260s\n",
            "Solving environment: - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\bdone\n",
            "\n",
            "\n",
            "==> WARNING: A newer version of conda exists. <==\n",
            "  current version: 4.4.10\n",
            "  latest version: 4.5.12\n",
            "\n",
            "Please update conda by running\n",
            "\n",
            "    $ conda update -n base conda\n",
            "\n",
            "\n",
            "\n",
            "## Package Plan ##\n",
            "\n",
            "  environment location: /usr/local\n",
            "\n",
            "  added / updated specs: \n",
            "    - pexpect\n",
            "\n",
            "\n",
            "The following packages will be UPDATED:\n",
            "\n",
            "    numpy-base:   1.14.3-py36h0ea5e3f_1                             --> 1.15.0-py36h3dfced4_0\n",
            "\n",
            "The following packages will be DOWNGRADED:\n",
            "\n",
            "    blas:         1.1-openblas                          conda-forge --> 1.0-mkl              \n",
            "    numpy:        1.15.4-py36_blas_openblashb06ca3d_0   conda-forge [blas_openblas] --> 1.15.0-py36h1b885b7_0\n",
            "    scikit-learn: 0.20.2-py36_blas_openblash00c3548_400 conda-forge [blas_openblas] --> 0.19.1-py36hedc7406_0\n",
            "    scipy:        1.2.0-py36_blas_openblashb06ca3d_200  conda-forge [blas_openblas] --> 1.1.0-py36hc49cb51_0 \n",
            "\n",
            "Preparing transaction: - \b\b\\ \b\b| \b\bdone\n",
            "Verifying transaction: - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\bdone\n",
            "Executing transaction: \\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\bdone\n",
            "\n",
            "real\t0m33.044s\n",
            "user\t0m31.369s\n",
            "sys\t0m1.298s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "06UTcNSAITTz",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## パッケージインストールが終わったらコードを実行してみましょう。\n",
        "- git clone でREINVENTを入手、テスト済みのデータをMishimasykのリポジトリから入手します。\n",
        "- 次いでREINVENT内のDataフォルダを丸ごと入れ替えます。\n",
        "- WGETやGITの部分は繰り返すと重複してフォルダが増えるので注意しましょう。!ls や!pwdで適当にチェックしてください。"
      ]
    },
    {
      "metadata": {
        "id": "lbeyp10e5fdD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "211dad9a-57de-4245-e925-f41117da665d"
      },
      "cell_type": "code",
      "source": [
        "import sys\n",
        "import os\n",
        "sys.path.append('/usr/local/lib/python3.6/site-packages/')\n",
        "from rdkit import Chem\n",
        "from rdkit import rdBase\n",
        "print(rdBase.rdkitVersion)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2018.09.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "27uy2hLv7ly4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1210
        },
        "outputId": "bd917aef-8dfe-46f3-8a91-a78ce9e39403"
      },
      "cell_type": "code",
      "source": [
        "!git clone https://github.com/MarcusOlivecrona/REINVENT.git\n",
        "os.chdir('./REINVENT')\n",
        "!wget https://github.com/Mishima-syk/13/raw/master/generator_hanson/data.zip\n",
        "!rm -r data/\n",
        "!unzip data.zip\n",
        "#Run generator!\n",
        "!time ./main.py --scoring-function tanimoto --scoring-function-kwargs query_structure 'COc1ccccc1' --num-steps 10"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "fatal: destination path 'REINVENT' already exists and is not an empty directory.\n",
            "--2019-01-03 05:06:57--  https://github.com/Mishima-syk/13/raw/master/generator_hanson/data.zip\n",
            "Resolving github.com (github.com)... 192.30.253.112, 192.30.253.113\n",
            "Connecting to github.com (github.com)|192.30.253.112|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://raw.githubusercontent.com/Mishima-syk/13/master/generator_hanson/data.zip [following]\n",
            "--2019-01-03 05:06:57--  https://raw.githubusercontent.com/Mishima-syk/13/master/generator_handson/data.zip\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 78454973 (75M) [application/zip]\n",
            "Saving to: ‘data.zip.1’\n",
            "\n",
            "data.zip.1          100%[===================>]  74.82M   185MB/s    in 0.4s    \n",
            "\n",
            "2019-01-03 05:06:58 (185 MB/s) - ‘data.zip.1’ saved [78454973/78454973]\n",
            "\n",
            "Archive:  data.zip\n",
            "   creating: data/\n",
            "  inflating: data/Prior.ckpt         \n",
            "   creating: data/results/\n",
            "   creating: data/results/run_2018-12-18-22_04_27/\n",
            "  inflating: data/results/run_2018-12-18-22_04_27/train_agent.py  \n",
            "  inflating: data/results/run_2018-12-18-22_04_27/sampled  \n",
            "  inflating: data/results/run_2018-12-18-22_04_27/memory  \n",
            "  inflating: data/results/run_2018-12-18-22_04_27/Agent.ckpt  \n",
            "   creating: data/results/run_2018-12-17-11_12_31/\n",
            "  inflating: data/results/run_2018-12-17-11_12_31/train_agent.py  \n",
            "  inflating: data/results/run_2018-12-17-11_12_31/sampled  \n",
            "  inflating: data/results/run_2018-12-17-11_12_31/memory  \n",
            "  inflating: data/results/run_2018-12-17-11_12_31/Agent.ckpt  \n",
            "  inflating: data/clf.pkl            \n",
            "  inflating: data/Voc                \n",
            "   creating: data/logs/\n",
            "  inflating: data/logs/weight_GRU_embedding.npy  \n",
            "  inflating: data/logs/init_weight_GRU_embedding.npy  \n",
            "  inflating: data/logs/weight_GRU_layer_2_w_hh.npy  \n",
            "  inflating: data/logs/init_weight_GRU_layer_2_w_hh.npy  \n",
            "  inflating: data/logs/init_weight_GRU_layer_2_w_ih.npy  \n",
            "  inflating: data/logs/weight_GRU_layer_2_w_ih.npy  \n",
            "  inflating: data/logs/Scores.npy    \n",
            "  inflating: data/logs/weight_GRU_layer_2_b_hh.npy  \n",
            "  inflating: data/logs/init_weight_GRU_layer_2_b_hh.npy  \n",
            "  inflating: data/logs/init_weight_GRU_layer_2_b_ih.npy  \n",
            "  inflating: data/logs/weight_GRU_layer_2_b_ih.npy  \n",
            "  inflating: data/logs/SMILES        \n",
            "  inflating: data/ChEMBL_filtered    \n",
            "   creating: data/reinvent_data/\n",
            "  inflating: data/reinvent_data/Prior.ckpt  \n",
            "  inflating: data/reinvent_data/log.txt  \n",
            "  inflating: data/reinvent_data/Voc  \n",
            "Traceback (most recent call last):\n",
            "  File \"./main.py\", line 59, in <module>\n",
            "    train_agent(**arg_dict)\n",
            "  File \"/content/REINVENT/train_agent.py\", line 38, in train_agent\n",
            "    Prior.rnn.load_state_dict(torch.load('data/Prior.ckpt'))\n",
            "  File \"/usr/local/lib/python3.6/site-packages/torch/nn/modules/module.py\", line 721, in load_state_dict\n",
            "    self.__class__.__name__, \"\\n\\t\".join(error_msgs)))\n",
            "RuntimeError: Error(s) in loading state_dict for MultiGRU:\n",
            "\tWhile copying the parameter named \"gru_1.bias_ih\", whose dimensions in the model are torch.Size([1536]) and whose dimensions in the checkpoint are torch.Size([1, 1536]).\n",
            "\tWhile copying the parameter named \"gru_1.bias_hh\", whose dimensions in the model are torch.Size([1536]) and whose dimensions in the checkpoint are torch.Size([1, 1536]).\n",
            "\tWhile copying the parameter named \"gru_2.bias_ih\", whose dimensions in the model are torch.Size([1536]) and whose dimensions in the checkpoint are torch.Size([1, 1536]).\n",
            "\tWhile copying the parameter named \"gru_2.bias_hh\", whose dimensions in the model are torch.Size([1536]) and whose dimensions in the checkpoint are torch.Size([1, 1536]).\n",
            "\tWhile copying the parameter named \"gru_3.bias_ih\", whose dimensions in the model are torch.Size([1536]) and whose dimensions in the checkpoint are torch.Size([1, 1536]).\n",
            "\tWhile copying the parameter named \"gru_3.bias_hh\", whose dimensions in the model are torch.Size([1536]) and whose dimensions in the checkpoint are torch.Size([1, 1536]).\n",
            "\n",
            "real\t0m6.621s\n",
            "user\t0m5.079s\n",
            "sys\t0m1.362s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "SNsDDPFLHKb8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 3828
        },
        "outputId": "c0c32775-8b2b-4893-c123-f01bb153257e"
      },
      "cell_type": "code",
      "source": [
        "!time ./main.py --scoring-function tanimoto --scoring-function-kwargs query_structure 'COc1ccccc1' --num-steps 10"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/site-packages/torch/nn/modules/module.py:482: UserWarning: src is not broadcastable to dst, but they have the same number of elements.  Falling back to deprecated pointwise behavior.\n",
            "  own_state[name].copy_(param)\n",
            "Model initialized, starting training...\n",
            "/content/REINVENT/model.py:99: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
            "  prob = F.softmax(logits)\n",
            "/content/REINVENT/model.py:100: UserWarning: Implicit dimension choice for log_softmax has been deprecated. Change the call to include dim=X as an argument.\n",
            "  log_prob = F.log_softmax(logits)\n",
            "/content/REINVENT/model.py:65: UserWarning: Implicit dimension choice for log_softmax has been deprecated. Change the call to include dim=X as an argument.\n",
            "  log_prob = F.log_softmax(logits)\n",
            "/content/REINVENT/model.py:66: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
            "  prob = F.softmax(logits)\n",
            "\n",
            "       Step 0   Fraction valid SMILES: 92.2  Time elapsed: 0.00h Time left: 0.01h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -19.30   -19.30  -17.07    0.11     COCc1n[nH]c(N2CCN(CC)CC2)n1\n",
            " -19.49   -19.49  -16.49    0.15     CCc1nn(CC(=O)OC)c(CC)c1Br\n",
            " -28.83   -28.83  -24.50    0.22     Cc1c(C(O)CN2CCC(c3ccccc3)C2)n(-c2ccccc2)nc1-c1ccco1\n",
            " -17.32   -17.32   -9.71    0.38     CCC1(c2ccccc2)OCCC(C)(C)NC(=O)N1\n",
            " -25.86   -25.86  -22.29    0.18     Clc1cc2c(cc1Cl)c(CNCc1ccncc1)cn2C1CCNCC1\n",
            " -31.43   -31.43  -25.71    0.29     N#CNC(=NC#N)NC(=Nc1ccc([N+](=O)[O-])cc1)c1ccc(O)cc1\n",
            " -41.71   -41.71  -36.81    0.25     COc1ccc(-c2nc(N(C(=O)c3ccc(Br)cc3Cl)S(=O)(=O)c3cccnc3Nc3ccccc3)sc2-c2cc(C)ccn2)cc1\n",
            " -29.71   -29.71  -24.10    0.28     COc1ccc(N(CC(=O)NC2CC3c4ccccc4CCc4cc5c(cc42)OCO5)CC3)cc1Cl\n",
            " -23.21   -23.21  -16.24    0.35     CCCN(CCC)CC(Cc1ccccc1)NC(=O)c1ccc(N(C)C)cc1\n",
            " -16.72   -16.72  -11.39    0.27     Cc1ccccc1NC(=O)CSc1nnc(-c2cccs2)n1Cc1ccccc1\n",
            "\n",
            "Best score in memory: 0.50\n",
            "\n",
            "       Step 1   Fraction valid SMILES: 90.6  Time elapsed: 0.00h Time left: 0.01h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -30.18   -29.79  -25.96    0.19     CCC1(CC)CC(CN2CC(N)C(c3ccc(F)cc3F)C2)OCC1\n",
            " -16.92   -16.32  -11.93    0.22     CCn1cc(NC(=O)NCc2ccncc2)cn1\n",
            " -35.31   -37.14  -32.50    0.23     Cc1cccc(OC(c2cccc(C=O)c2)C(O)C(=O)N2CC(C)N(C(=O)OC(C)(C)C)CC2C(=O)NC(C)(C)C)c1\n",
            " -31.34   -32.72  -28.44    0.21     CN(CC(=O)Nc1c(C(=O)NN=Cc2cccc(Br)c2)cnn1C)C(=O)c1cccc(F)c1\n",
            " -21.84   -21.70  -13.76    0.40     COc1ccc2c(c1)OCC(CCOc1ccc(C(C)C)cc1)C2\n",
            " -19.30   -20.29  -14.58    0.29     Cc1nn(-c2cccc(C(N)=O)c2)c2c1c(C)nn2C\n",
            " -53.11   -54.86  -54.86    0.00     CC(C)(C)c1cc(-c2nc3c(c4c2CC(c2ccccc2)OC4)CN(Cc2c(O)ccc4ccccc24)CC3)c(C)c(C)n1CCN(C)C\n",
            " -19.22   -18.84  -13.43    0.27     CCOC(=O)CCCn1cc(C=C(C#N)C#N)c2ccccc21\n",
            " -23.59   -23.62  -16.48    0.36     CCN1C(=O)NC(=Nc2ccc(C(F)(F)F)cc2)c2ccccc21\n",
            " -20.54   -20.58  -20.01    0.03     CC1(C(O)(CO)CO)CCC2(C)CCC3(C)C4=CC(=O)C=CC4(C)C3CCC21C\n",
            "\n",
            "Best score in memory: 0.56\n",
            "\n",
            "       Step 2   Fraction valid SMILES: 92.2  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -25.21   -25.32  -14.21    0.56     CC1(C)C(N)CC(O)(COc2ccccc2)C(=O)O1\n",
            " -27.67   -28.39  -24.38    0.20     O=C(CC1SC(c2ccco2)N(NC(=O)c2cc3ccccc3[nH]2)C1=O)NCc1cccnc1\n",
            " -18.05   -17.47  -12.81    0.23     O=C(CNC(=O)c1cccc(C(F)(F)F)c1)NC1CCN(CC=Cc2ccccc2)CC1\n",
            " -19.92   -22.72  -16.89    0.29     COc1ccc(NC(=O)Cc2c(C(=O)O)n(Cc3ccccc3Cl)c3ccc(Cl)cc23)cc1Cl\n",
            " -23.85   -23.34  -16.78    0.33     CN(C)CCC=C1c2cc(Br)ccc2OC(=O)C1C\n",
            " -34.47   -36.14  -32.54    0.18     COC(=O)C1CCN(C(=O)c2ccc(C(=O)C(=O)Nc3cc(C)cc(C(F)(F)F)c3)cc2C(F)(F)F)CC1\n",
            " -25.53   -26.39  -20.53    0.29     Cc1cc(OCc2ccc(S(=O)(=O)Cc3cccc(C#N)c3)cc2)ccn1\n",
            " -14.76   -14.38   -8.92    0.27     Cc1ccc(-c2csc3nc4n(c(=O)c23)CCC4)cc1\n",
            " -32.65   -33.74  -27.86    0.29     COC(=O)N(C)CCCOc1ccc2c(c1)c(-c1ccc(N)cc1)nc(=O)n2Cc1ccccc1\n",
            " -24.83   -24.47  -18.70    0.29     CC(C)(C)OCC(O)(Cn1cncn1)c1ccc(OCc2cccc(F)c2)cc1Cl\n",
            "\n",
            "Best score in memory: 0.56\n",
            "\n",
            "       Step 3   Fraction valid SMILES: 93.8  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -14.91   -15.25  -14.20    0.05     CC(N)C(=O)NC(C=O)CC(C)C\n",
            " -20.83   -21.51  -15.23    0.31     Cc1ccc(COc2cccnc2Oc2ccc(C(=O)c3ccccc3)cc2)cc1\n",
            " -23.45   -23.13  -17.58    0.28     N=C(CSC(=O)N1CCOCC1)Nc1cccc(Cl)c1\n",
            " -18.15   -16.77  -12.63    0.21     CC(C)NC(=O)Nc1cc2c3c(c1)CC(=O)N3CCC2\n",
            " -20.24   -20.20  -14.94    0.26     CC1COc2ccc(C#CC3(O)CCN(C)CC3O)cc2NC1=O\n",
            " -28.88   -30.72  -26.13    0.23     CC(C)C1CC(=O)N(c2ccc3c(c2)OCC3)C(=O)N1CC(=O)N(C)C\n",
            " -25.24   -27.32  -22.37    0.25     Cc1ccc(C(=O)N2c3cccc(O)c3N=C(c3ccccc3)C2c2ccc(Br)cc2)s1\n",
            " -21.82   -22.36  -17.11    0.26     O=C(NO)C1(S(=O)(=O)c2ccc(Oc3ccc(Cl)c(Cl)c3)c(Cl)c2)CCNCC1\n",
            " -27.21   -27.94  -22.56    0.27     CC(CCc1ccccc1)Nc1nc(C(=O)Nc2ncns2)ccc1C#N\n",
            " -20.58   -21.18  -16.57    0.23     O=C(NC(Cc1ccccc1)C(=O)O)c1cccc(NS(=O)(=O)c2cccc([N+](=O)[O-])c2)c1\n",
            "\n",
            "Best score in memory: 0.56\n",
            "\n",
            "       Step 4   Fraction valid SMILES: 90.6  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -19.85   -22.67  -16.76    0.30     COc1ccc(C2C(Cl)C(=O)N2NC(=O)c2ccc(-n3cncn3)cc2)cc1O\n",
            " -16.12   -15.59   -8.71    0.34     CCOc1cc(C=NNC(N)=O)ccc1O\n",
            " -18.51   -20.51  -15.03    0.27     COc1cc(N)c(Cl)cc1C(=O)C1CCCN1C(=O)OC(C)(C)C\n",
            " -31.32   -34.31  -30.00    0.22     Cc1c(Nc2c(C(N)=O)cnn2C)cccc1C(=O)NCC(=O)Nc1ccncc1\n",
            " -17.58   -20.07  -14.05    0.30     O=c1[nH]c(=O)n(COCCO)cc1-c1cc(-c2ccccc2)no1\n",
            " -21.48   -21.08  -18.37    0.14     Cc1cc(C(=O)NCC(C)(C)CN2CCN(C)CC2)on1\n",
            " -32.27   -35.78  -32.74    0.15     O=C(c1ccc(-c2ccnc3c(C(F)(F)F)cc(N4CCOCC4)cc23)cc1)N1C2CCC1CC(O)(c1cccc(C(F)(F)F)c1)C2\n",
            " -17.81   -18.28  -12.04    0.31     COc1ccc2nc(O)c(CN(Cc3cccnc3)C3CCS(=O)(=O)C3)cc2c1\n",
            " -26.57   -26.02  -20.18    0.29     Cc1cccc(-c2ccccc2COc2ccc(S(=O)(=O)Nc3cscn3)cc2C)n1\n",
            " -18.51   -18.22  -10.80    0.37     COc1ccc(OC)c(-c2ccc(C(=O)NC(CCC(=O)O)C(=O)O)s2)c1\n",
            "\n",
            "Best score in memory: 0.56\n",
            "\n",
            "       Step 5   Fraction valid SMILES: 87.5  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -43.28   -45.85  -42.41    0.17     CS(=O)(=O)Nc1ccc(C(F)(F)F)cc1N=Nc1c(O)c(-c2ccccc2)c2[n+]3cc(F)ccc3n(CCCc3ccccc3)c2c1F\n",
            " -14.60   -15.73   -8.99    0.34     COc1cc2cc[n+](CCc3ccc(C)cc3)cc2cc1OC\n",
            " -22.53   -22.22  -14.32    0.39     COc1ccccc1C1=Nc2nc3ccccn3c2C(=O)C(C(C)=O)=C1\n",
            " -34.70   -40.86  -40.40    0.02     CC(=O)OC(CC=C(C)C)CC(C)C1CC(=O)C2(C)C3=C(C(=O)CC12)C1(C)C(=O)C=CC(O)C(C)(C)C1CC3\n",
            " -21.30   -21.61  -14.47    0.36     O=C(NCCOc1ccccc1F)C(=O)NCCCc1ccccc1\n",
            " -19.24   -19.77  -12.11    0.38     COc1cc(-c2nnc3ccc(Sc4ccccc4)cn23)cc(OC)c1OC\n",
            " -24.51   -25.38  -21.30    0.20     O=C(Nc1ccc(F)cc1F)c1cc(-c2ccc(S(=O)(=O)N3CCc4ccccc43)cc2)no1\n",
            " -26.12   -26.75  -22.96    0.19     CC(C)c1cc(C(=O)N2CCC(c3nc(-c4ccc(C#N)cc4)no3)CC2)ccc1Cl\n",
            " -22.80   -21.67  -16.73    0.25     Cc1cc(F)ccc1C(=O)N(Cc1cccc(F)c1)C1CCS(=O)(=O)C1\n",
            " -32.40   -35.03  -31.31    0.19     O=C(CN(C(=O)Cc1ccc(Cl)cc1)c1nc2cc(Cl)c(F)cc2s1)c1ccc(Cl)cc1\n",
            "\n",
            "Best score in memory: 0.56\n",
            "\n",
            "       Step 6   Fraction valid SMILES: 95.3  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -19.50   -18.66  -12.32    0.32     COC(=O)C(Cc1cccc([N+](=O)[O-])c1)NC(=O)CN\n",
            " -28.99   -28.76  -23.40    0.27     CC#CCOCC(C)Oc1ccc(S(=O)(=O)CC2(C(=O)NO)CCN(C(=O)OC)CC2)cc1\n",
            " -18.81   -17.92  -12.63    0.26     O=C(c1ccco1)N1N=C(c2cccnc2)CC1c1ccccc1F\n",
            " -32.89   -35.35  -31.54    0.19     COC(=O)C1=C(C(=O)OC)C2Nc3c(C)c(Br)ccc3NC2N(Cc2ccc([N+](=O)[O-])cc2)C1=O\n",
            " -14.86   -16.35  -12.13    0.21     O=C(COC(=O)c1ccc(Cl)nc1)NCCO\n",
            " -16.83   -15.85  -10.83    0.25     CC(=O)Nc1ccc(NC(=O)c2cc(-c3ccccn3)nc3ccccc23)cc1\n",
            " -23.10   -21.96  -16.15    0.29     NC(C(=O)NO)C(=O)NCCc1cccc([N+](=O)[O-])c1\n",
            " -20.34   -20.73  -14.14    0.33     CCOc1ccc(C=NNc2nc(-c3ccccc3)cnc2-c2ccccc2)cc1\n",
            " -22.88   -26.17  -21.13    0.25     Cc1ccc2cccc(S(=O)(=O)Nc3cc(-c4cccc(Cl)c4)on3)c2c1\n",
            " -62.45   -71.55  -71.55    0.00     Cc1ccc(C)c(OCC(=O)NC(N=Nc2ccc(S(=O)(=O)Nc3ccc4onc(N)c4c3)c3cc(C)cc(C(=N)N)c3)cc2)c1CNC(=O)CCN=C(N)N\n",
            "\n",
            "Best score in memory: 0.57\n",
            "\n",
            "       Step 7   Fraction valid SMILES: 96.9  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -38.07   -40.48  -35.86    0.23     CC1CCN(c2cc(CNC(=O)c3ccccc3[N+](=O)[O-])cc3oc(C(=O)N(C)C)cc23)C1\n",
            " -24.01   -25.57  -18.15    0.37     COc1ccccc1-c1nc(SCc2ccc(Br)cc2)nc(C)c1C#N\n",
            " -26.98   -30.61  -25.80    0.24     Cc1sc(Cc2ccc(C(=O)NC3CCS(=O)(=O)CC3)cc2C)nc1-c1ccccc1\n",
            " -34.95   -34.12  -29.92    0.21     Nc1cccc2c1C(=O)N(C1CCN(CC=Cc3ccccc3)CN(Cc3ccccc3)C1=O)S2\n",
            " -16.71   -20.54  -12.59    0.40     COc1cccc(OC)c1Oc1ccccc1CN(C)Cc1ccccc1\n",
            " -22.06   -31.93  -26.31    0.28     COc1cc2c(cc1O)C(C)(c1ccccc1)Oc1ccc(cc1)Oc1cccc(c1)Oc1ccccc1OC2\n",
            " -22.71   -21.76  -15.85    0.30     O=C(COc1ccccc1C(=O)N1CCOCC1)Nc1ccc(C(F)(F)F)cc1\n",
            " -23.57   -22.49  -13.92    0.43     C[n+]1cn(C)nc1Sc1ccc(F)cc1\n",
            " -23.38   -22.24  -16.10    0.31     CCOc1ccc(CCNC(=O)C2CCN(C(=O)N3Cc4ccccc4C3)CC2)cc1\n",
            " -27.20   -29.29  -29.29    0.00     Cc1c(-c2ccccc2)cc(-c2ccccc2)nn1CC(=O)NC(C)c1ccccc1)c1ccccc1\n",
            "\n",
            "Best score in memory: 0.57\n",
            "\n",
            "       Step 8   Fraction valid SMILES: 93.8  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -25.03   -24.69  -21.41    0.16     COC1CC(CC#Cc2ccc(N3CCOCC3)nc2)C(O)C(O)C1O\n",
            " -24.68   -24.17  -19.06    0.26     O=[N+]([O-])c1cccc(CSc2nc(-c3ccc(F)cc3)c(C(=O)c3ccccc3)s2)c1\n",
            " -11.98   -12.71   -5.57    0.36     O=c1cc(-c2ccccc2)oc2cccc(O)c12\n",
            " -15.92   -18.17  -11.45    0.34     O=C(O)c1cn(S(=O)(=O)c2ccccc2Cl)c2ccccc12\n",
            " -22.37   -22.50  -16.59    0.30     O=C(Nc1ccc([N+](=O)[O-])cc1)c1ccc(OCc2ccc([N+](=O)[O-])cc2)cc1\n",
            " -15.64   -17.18   -9.71    0.37     COc1cc(-n2cnc3cc(Cl)ccc3c2=O)ccc1O\n",
            " -16.30   -19.17  -11.52    0.38     COc1cc(C=CC(=O)O)ccc1OCc1cccc(C(=O)O)n1\n",
            " -30.08   -35.79  -30.40    0.27     COc1cc(C(N)=NO)ccc1-c1cncc2cc(-c3cccc4ccccc34)c(OC(C)C)nc12\n",
            " -16.78   -19.70  -12.65    0.35     COc1cc(OC)cc(C(=O)N(Cc2ccccc2Br)C2CCS(=O)(=O)C2)c1\n",
            " -18.54   -18.09  -10.09    0.40     COc1cccc(CN2CCC3C2CCC(=O)N3c2ccccn2)c1\n",
            "\n",
            "Best score in memory: 0.77\n",
            "\n",
            "       Step 9   Fraction valid SMILES: 92.2  Time elapsed: 0.00h Time left: 0.00h\n",
            "  Agent    Prior   Target   Score             SMILES\n",
            " -19.02   -19.23  -13.44    0.29     O=C(Cc1sc2ccccc2c1Cl)NCCc1ccc(Cl)cc1\n",
            " -13.45   -16.52   -7.41    0.46     COc1ccc(OC)c(C(O)c2nc(-c3ccccc3Cl)cs2)c1\n",
            " -15.05   -14.57   -6.09    0.42     COc1ccc2cc3cc(C(=O)NC4CC4)sc3nc2c1\n",
            " -21.91   -22.32  -15.27    0.35     COc1ccc2nc(Sc3nncn3Cc3ccc(Cl)cc3)sc2c1\n",
            " -40.86   -46.28  -46.28    0.00     Cc1cc(Nc2c(-c3cccc(NC(=O)COc4cccc(CN5CCOCC5)c4)c3)cc3ccccc3n2)c(C)nn1C\n",
            " -15.03   -16.21   -8.59    0.38     CC(=O)NC(Cc1cccc2ccccc12)C(=O)O\n",
            " -32.29   -35.94  -31.61    0.22     Cc1nc2c(ccc3[nH]ccc32)c2c(-c3ccccc3)cc(-c3cccc(O)c3)nc12\n",
            " -25.34   -26.23  -16.87    0.47     COc1ccc(C(C)C(C)C2COC(=O)C2(C)O)cc1\n",
            " -19.16   -17.83  -10.61    0.36     COc1ccccc1COCCCOc1ccc(N2C(=O)CNCC2C(C)N)cc1\n",
            " -17.17   -19.02  -13.30    0.29     Cc1ccc(N2C(=O)NC(=O)C(=CN(C)C)c3ccccc32)cc1\n",
            "\n",
            "********************************************************************************\n",
            "\n",
            "         Best recorded SMILES: \n",
            "\n",
            "Score     Prior log P     SMILES\n",
            "\n",
            "0.77   -17.10        COc1cccc(OCCO)c1\n",
            "0.57   -17.23        COc1cccc(C2(O)OC(=O)c3ccccc32)c1\n",
            "0.56   -25.32        CC1(C)C(N)CC(O)(COc2ccccc2)C(=O)O1\n",
            "0.54   -18.29        COc1ccc(C(C)=NOC(=O)N(C)C)cc1\n",
            "0.51   -18.49        COc1cccc(OCC(O)CNCCN(C)C)c1\n",
            "0.51   -16.16        O=C(Nc1ccc(OCc2ccccc2)cc1)C(Cl)Cl\n",
            "0.50   -18.00        COc1cc(C(=O)O)ccc1NCc1ccccc1\n",
            "0.50   -16.86        COc1ccc2c(c1)CCc1cc(N(C)C)ccc1-2\n",
            "0.50   -16.55        [N-]=[N+]=NCc1ccccc1F\n",
            "0.50   -18.19        COc1ccc(C2Nc3ccccc3-n3cnnc32)cc1\n",
            "0.50   -19.68        COc1ccc2c(c1)C(=O)C(O)=C(Br)C2=O\n",
            "0.49   -15.37        Clc1ccccc1OCCOCCOc1ccccc1\n",
            "0.48   -24.38        COc1ccccc1N1CC(C(=O)O)C(c2nccs2)C1\n",
            "0.48   -18.20        COc1ccc(NCc2ccccc2)c(O)n1\n",
            "0.48   -22.01        COc1cc2c(cc1C(=O)O)Oc1ccccc1CC2\n",
            "0.47   -20.32        COc1ccc(OCC(=O)c2cc(C)cc3ccccc23)cc1\n",
            "0.47   -15.55        COc1ccccc1NC(=O)c1ccc2[nH]cnc2c1\n",
            "0.47   -26.23        COc1ccc(C(C)C(C)C2COC(=O)C2(C)O)cc1\n",
            "0.47   -18.77        COc1ccc(OCCON=C(N)N)cc1OC\n",
            "0.47   -20.17        COC(O)=C1C(C)=Nc2ccccc2N=C1NC\n",
            "0.47   -14.08        COc1ccc(OC)c2c1CC(N)CC2\n",
            "0.46   -16.65        COc1ccc(CCNCc2cccc(-c3ccccc3)c2)cc1\n",
            "0.46   -17.95        COc1cccc(OC2CCN(C(=O)NC(C)C)CC2)c1\n",
            "0.46   -14.38        COc1ccc(NC(=O)c2ccco2)cc1OC\n",
            "0.46   -24.03        Cc1cc(Cl)ccc1OCc1ccccc1C=O\n",
            "0.46   -22.31        COc1c(C=CC(=O)O)cccc1OCCF\n",
            "0.46   -16.52        COc1ccc(OC)c(C(O)c2nc(-c3ccccc3Cl)cs2)c1\n",
            "0.45   -21.03        COc1ccccc1-c1cncc(C#N)c1-c1ccccc1\n",
            "0.45   -16.64        COc1ccc(C(=O)O)c(Nc2cccc(C(F)(F)F)c2)c1\n",
            "0.45   -17.31        CCOc1ccc(NC(S)=Nc2nccs2)cc1\n",
            "0.45   -12.16        Nc1cccc(C(=O)O)c1\n",
            "0.45   -13.13        CNC(=O)C(=NOC)c1ccccc1COc1ccc(Cl)cc1\n",
            "0.45   -16.24        Cc1ccc(OCC(=O)NCC(c2ccccc2)N(C)C)cc1\n",
            "0.45   -16.46        COc1ccccc1CNC(=O)COC(=O)c1ccccc1Cl\n",
            "0.45   -23.32        C=C(c1ccccc1)C(CO)C(C)(C)C\n",
            "0.44   -15.85        CC(=O)Nc1ccc(SCc2ccc(Cl)cc2)cc1\n",
            "0.43   -21.83        O=C(Oc1ccccc1)c1ccccc1Oc1ccccc1F\n",
            "0.43   -25.12        CCOc1cccc(CCN(C)C(=S)OCC)c1\n",
            "0.43   -23.03        COc1ccccc1NC(=O)N(C)c1ccc2cc[nH]c2c1\n",
            "0.43   -15.45        COc1ccc(C=C2NC(=S)N(C(Cc3ccccc3)C(=O)O)C2=O)cc1\n",
            "0.43   -17.95        COc1ccc(OC)c2c1OC(C)(C)C(=O)C2\n",
            "0.43   -15.69        COc1ccc(S(=O)(=O)CCC(=O)N2CCOCC2)cc1\n",
            "0.43   -19.55        O=C(Oc1ccncc1)c1sc2ccccc2c1Cl\n",
            "0.43   -22.49        C[n+]1cn(C)nc1Sc1ccc(F)cc1\n",
            "0.43   -19.63        CCOC(=O)COc1ccc2c(-c3ccccc3)c[nH]c2c1\n",
            "0.43   -17.67        COc1ccc(C(=O)OCC(=O)c2ccc([N+](=O)[O-])cc2)cc1\n",
            "0.42   -14.57        COc1ccc2cc3cc(C(=O)NC4CC4)sc3nc2c1\n",
            "0.42   -15.93        COc1ccc(C(C)(O)c2nc(-c3ccc(O)cc3)cs2)cc1\n",
            "0.42   -25.67        COc1ccc2c(c1)CCC1OCC=C2C(O)O1\n",
            "0.42   -25.47        Cc1cc(Nc2ccccc2)cc(CSc2ccc(C#N)cc2)n1\n",
            "\n",
            "********************************************************************************\n",
            "\n",
            "\n",
            "real\t0m15.913s\n",
            "user\t0m12.237s\n",
            "sys\t0m3.577s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "jeFNI0L6N863",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## 今度は学習済みのモデルからダイレクトにサンプリングします。\n",
        "- vocには学習データから作成した語彙（Smilesに使われるキャラクター）が含まれます。\n",
        "- Agent objectを作成後学習済みの重みをロードします。\n",
        "- ロード部分をスキップしてSamplingするとまともな構造は出ません。確認してみましょう。"
      ]
    },
    {
      "metadata": {
        "id": "h1fkXTM_IAh1",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "\n",
        "import os\n",
        "from model import RNN\n",
        "from data_structs import Vocabulary, Experience\n",
        "from utils import Variable, seq_to_smiles, fraction_valid_smiles, unique\n",
        "from rdkit.Chem.Draw import IPythonConsole\n",
        "from rdkit.Chem import Draw\n",
        "voc = Vocabulary(init_from_file=\"data/Voc\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "a1bR799KKSlx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 71
        },
        "outputId": "4cd4fbdc-8919-4315-bf1c-5486ac33069b"
      },
      "cell_type": "code",
      "source": [
        "Agent = RNN(voc)\n",
        "Agent.rnn.load_state_dict(torch.load('data/Prior.ckpt', map_location=lambda storage, loc: storage))"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/site-packages/torch/nn/modules/module.py:482: UserWarning: src is not broadcastable to dst, but they have the same number of elements.  Falling back to deprecated pointwise behavior.\n",
            "  own_state[name].copy_(param)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "eYqW-JwtK9k2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        },
        "outputId": "60554d69-e135-4028-d436-eb2a5b890831"
      },
      "cell_type": "code",
      "source": [
        "seqs, agent_likelihood, entropy = Agent.sample(20)"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/REINVENT/model.py:99: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
            "  prob = F.softmax(logits)\n",
            "/content/REINVENT/model.py:100: UserWarning: Implicit dimension choice for log_softmax has been deprecated. Change the call to include dim=X as an argument.\n",
            "  log_prob = F.log_softmax(logits)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "hQkBDY4ZLhCg",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "smiles = seq_to_smiles(seqs, voc)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "5729QvDMLjDo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 340
        },
        "outputId": "c90ae676-4836-4ce8-99da-a889a14c2cbe"
      },
      "cell_type": "code",
      "source": [
        "validmols = []\n",
        "for smi in smiles:\n",
        "  mol = Chem.MolFromSmiles(smi)\n",
        "  if mol != None:\n",
        "    print(smi)\n",
        "    validmols.append(mol)"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "CC1COCCN1c1nc(N2CCOCC2)nc(C2CCNC2)n1\n",
            "Cc1nc(NCCCN2CCN(C)CC2)c2cnn(-c3cccc(Cl)c3)c2n1\n",
            "O=C(Nc1cccc2cccnc12)c1ccc(NCc2cccc(Cl)c2)cc1\n",
            "CC(C)(C)c1cc(C(=O)Nc2cc3c(c(OC(=O)N4CCOCC4)c(=O)[nH]3)c(F)cn2)c(F)cn1\n",
            "COc1ccc(-n2sc(=NC(=O)c3ccc(OC)c(OC)c3)nn2)cc1\n",
            "C=CCOC(=O)NCCC[N+](C)(C)CCSC\n",
            "CN=C1N=C(Nc2nccs2)SC1=Nc1ccccc1\n",
            "Cc1ccc(NC(S)=NNC(=S)NC(=O)C=Cc2ccccc2)cc1C\n",
            "CCOC(=O)c1cc2c3ccccc3n(Cc3ccccc3)c2c[n+]1CC\n",
            "COc1ccc2c(c1)NC1(CCN(Cc3nccs3)CC1)CN2\n",
            "COC(=O)C1=C(c2ccc(F)cc2)S(=O)(=O)N=C1N1CCN(C(=O)c2cccc(F)c2)CC1\n",
            "Cc1cn2cc(C(F)(F)F)nc2c(C#Cc2ccccn2)n1\n",
            "CCN(CC)c1ccc(C=NN=C(N)NO)cc1\n",
            "CCC(=O)NCCc1cc(Br)ccc1OCc1ccccc1C#N\n",
            "Cc1ccc(Cl)c2c1CC(O)C(NCC1(c3ccoc3)CCC1)C2\n",
            "CCn1nc2c(c1C)C(=O)C(=Cc1ccncc1)CC2\n",
            "Cc1noc(NC(=O)c2ccccc2)c1NC(C)=C1C(=O)CC(c2ccccc2)CC1=O\n",
            "O=C(Nc1nnc(Cc2ccc(Cl)cc2O)s1)c1ccccc1\n",
            "Cc1ccccc1-c1cc2ccccc2nc1SCC(=O)NCC1CCCO1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "yNdbhEGjLooW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "edde075f-865e-431f-803d-924a866a1c7b"
      },
      "cell_type": "code",
      "source": [
        "len(validmols)"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "19"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "metadata": {
        "id": "mEgTpAvQLvvz",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 817
        },
        "outputId": "5acef38c-475c-452c-e051-7cdc85caae0f"
      },
      "cell_type": "code",
      "source": [
        "Draw.MolsToGridImage(validmols, molsPerRow=5)"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA+gAAAMgCAIAAAA/et9qAAEAAElEQVR4nOzdd1hT2RIA8ElC6NIU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\nrkJggg==\n",
            "text/plain": [
              "<PIL.PngImagePlugin.PngImageFile image mode=RGB size=1000x800 at 0x7F556A00BFD0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 52
        }
      ]
    },
    {
      "metadata": {
        "id": "IEXSTrB7MPyc",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "\n"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}