simplified_keras_cnn_gcn_rnn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.03356,
     "end_time": "2020-10-22T19:24:05.271404",
     "exception": false,
     "start_time": "2020-10-22T19:24:05.237844",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:05.341030Z",
     "iopub.status.busy": "2020-10-22T19:24:05.340014Z",
     "iopub.status.idle": "2020-10-22T19:24:12.242018Z",
     "shell.execute_reply": "2020-10-22T19:24:12.241121Z"
    },
    "papermill": {
     "duration": 6.93986,
     "end_time": "2020-10-22T19:24:12.242144",
     "exception": false,
     "start_time": "2020-10-22T19:24:05.302284",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import json\n",
    "import os\n",
    "from tqdm import tqdm\n",
    "\n",
    "# sklearn\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# tensorflow and keras\n",
    "from keras.utils.vis_utils import plot_model\n",
    "\n",
    "import tensorflow.keras.layers as L\n",
    "import keras.backend as K\n",
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:12.322918Z",
     "iopub.status.busy": "2020-10-22T19:24:12.322060Z",
     "iopub.status.idle": "2020-10-22T19:24:23.660408Z",
     "shell.execute_reply": "2020-10-22T19:24:23.660990Z"
    },
    "papermill": {
     "duration": 11.382788,
     "end_time": "2020-10-22T19:24:23.661176",
     "exception": false,
     "start_time": "2020-10-22T19:24:12.278388",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "!pip install spektral -q"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:23.727745Z",
     "iopub.status.busy": "2020-10-22T19:24:23.726646Z",
     "iopub.status.idle": "2020-10-22T19:24:24.327123Z",
     "shell.execute_reply": "2020-10-22T19:24:24.326462Z"
    },
    "papermill": {
     "duration": 0.635759,
     "end_time": "2020-10-22T19:24:24.327257",
     "exception": false,
     "start_time": "2020-10-22T19:24:23.691498",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from spektral.layers import GraphConv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.031492,
     "end_time": "2020-10-22T19:24:24.396374",
     "exception": false,
     "start_time": "2020-10-22T19:24:24.364882",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Read Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:24.467335Z",
     "iopub.status.busy": "2020-10-22T19:24:24.466383Z",
     "iopub.status.idle": "2020-10-22T19:24:25.097686Z",
     "shell.execute_reply": "2020-10-22T19:24:25.096972Z"
    },
    "papermill": {
     "duration": 0.670415,
     "end_time": "2020-10-22T19:24:25.097828",
     "exception": false,
     "start_time": "2020-10-22T19:24:24.427413",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_json_path = \"/kaggle/input/stanford-covid-vaccine/train.json\"\n",
    "test_json_path = \"/kaggle/input/stanford-covid-vaccine/test.json\"\n",
    "sample_sub_path = \"/kaggle/input/stanford-covid-vaccine/sample_submission.csv\"\n",
    "\n",
    "output_path = \"./\"\n",
    "bpps_path = \"/kaggle/input/stanford-covid-vaccine/bpps\"\n",
    "\n",
    "train_df = pd.read_json(train_json_path, lines=True)\n",
    "test_df = pd.read_json(test_json_path, lines=True)\n",
    "\n",
    "# there are 2 part of the test set, they have different seq length\n",
    "public_df = test_df.query(\"seq_length == 107\").copy()\n",
    "private_df = test_df.query(\"seq_length == 130\").copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:25.167988Z",
     "iopub.status.busy": "2020-10-22T19:24:25.166893Z",
     "iopub.status.idle": "2020-10-22T19:24:25.171428Z",
     "shell.execute_reply": "2020-10-22T19:24:25.171960Z"
    },
    "papermill": {
     "duration": 0.043134,
     "end_time": "2020-10-22T19:24:25.172100",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.128966",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2400, 19)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.030743,
     "end_time": "2020-10-22T19:24:25.234657",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.203914",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Prediction Columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:25.332820Z",
     "iopub.status.busy": "2020-10-22T19:24:25.317121Z",
     "iopub.status.idle": "2020-10-22T19:24:25.363765Z",
     "shell.execute_reply": "2020-10-22T19:24:25.364323Z"
    },
    "papermill": {
     "duration": 0.097259,
     "end_time": "2020-10-22T19:24:25.364467",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.267208",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>reactivity</th>\n",
       "      <th>deg_Mg_pH10</th>\n",
       "      <th>deg_pH10</th>\n",
       "      <th>deg_Mg_50C</th>\n",
       "      <th>deg_50C</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>[0.3297, 1.5693000000000001, 1.1227, 0.8686, 0...</td>\n",
       "      <td>[0.7556, 2.983, 0.2526, 1.3789, 0.637600000000...</td>\n",
       "      <td>[2.3375, 3.5060000000000002, 0.3008, 1.0108, 0...</td>\n",
       "      <td>[0.35810000000000003, 2.9683, 0.2589, 1.4552, ...</td>\n",
       "      <td>[0.6382, 3.4773, 0.9988, 1.3228, 0.78770000000...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>[0.0, 0.0, 0.0, 2.2965, 0.0, 0.0, 0.0, 0.0, 0....</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[4.947, 4.4523, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[4.8511, 4.0426, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[7.6692, 0.0, 10.9561, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>[0.44820000000000004, 1.4822, 1.1819, 0.743400...</td>\n",
       "      <td>[0.2504, 1.4021, 0.9804, 0.49670000000000003, ...</td>\n",
       "      <td>[2.243, 2.9361, 1.0553, 0.721, 0.6396000000000...</td>\n",
       "      <td>[0.5163, 1.6823000000000001, 1.0426, 0.7902, 0...</td>\n",
       "      <td>[0.9501000000000001, 1.7974999999999999, 1.499...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>[0.0, 2.2399, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0....</td>\n",
       "      <td>[0.0, -0.5083, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0...</td>\n",
       "      <td>[3.4248, 6.8128, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[0.0, -0.8365, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0...</td>\n",
       "      <td>[7.6692, -1.3223, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>[0.8267, 2.6577, 2.8481, 0.40090000000000003, ...</td>\n",
       "      <td>[2.1058, 3.138, 2.5437000000000003, 1.0932, 0....</td>\n",
       "      <td>[4.7366, 4.6243, 1.2068, 1.1538, 0.0, 0.0, 0.7...</td>\n",
       "      <td>[2.2052, 1.7947000000000002, 0.7457, 3.1233, 0...</td>\n",
       "      <td>[0.0, 5.1198, -0.3551, -0.3518, 0.0, 0.0, 0.0,...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          reactivity  \\\n",
       "0  [0.3297, 1.5693000000000001, 1.1227, 0.8686, 0...   \n",
       "1  [0.0, 0.0, 0.0, 2.2965, 0.0, 0.0, 0.0, 0.0, 0....   \n",
       "2  [0.44820000000000004, 1.4822, 1.1819, 0.743400...   \n",
       "3  [0.0, 2.2399, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0....   \n",
       "4  [0.8267, 2.6577, 2.8481, 0.40090000000000003, ...   \n",
       "\n",
       "                                         deg_Mg_pH10  \\\n",
       "0  [0.7556, 2.983, 0.2526, 1.3789, 0.637600000000...   \n",
       "1  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "2  [0.2504, 1.4021, 0.9804, 0.49670000000000003, ...   \n",
       "3  [0.0, -0.5083, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0...   \n",
       "4  [2.1058, 3.138, 2.5437000000000003, 1.0932, 0....   \n",
       "\n",
       "                                            deg_pH10  \\\n",
       "0  [2.3375, 3.5060000000000002, 0.3008, 1.0108, 0...   \n",
       "1  [4.947, 4.4523, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "2  [2.243, 2.9361, 1.0553, 0.721, 0.6396000000000...   \n",
       "3  [3.4248, 6.8128, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "4  [4.7366, 4.6243, 1.2068, 1.1538, 0.0, 0.0, 0.7...   \n",
       "\n",
       "                                          deg_Mg_50C  \\\n",
       "0  [0.35810000000000003, 2.9683, 0.2589, 1.4552, ...   \n",
       "1  [4.8511, 4.0426, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "2  [0.5163, 1.6823000000000001, 1.0426, 0.7902, 0...   \n",
       "3  [0.0, -0.8365, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0...   \n",
       "4  [2.2052, 1.7947000000000002, 0.7457, 3.1233, 0...   \n",
       "\n",
       "                                             deg_50C  \n",
       "0  [0.6382, 3.4773, 0.9988, 1.3228, 0.78770000000...  \n",
       "1  [7.6692, 0.0, 10.9561, 0.0, 0.0, 0.0, 0.0, 0.0...  \n",
       "2  [0.9501000000000001, 1.7974999999999999, 1.499...  \n",
       "3  [7.6692, -1.3223, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...  \n",
       "4  [0.0, 5.1198, -0.3551, -0.3518, 0.0, 0.0, 0.0,...  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_cols = ['reactivity', 'deg_Mg_pH10', 'deg_pH10', 'deg_Mg_50C', 'deg_50C']\n",
    "train_df[pred_cols].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:25.550634Z",
     "iopub.status.busy": "2020-10-22T19:24:25.549579Z",
     "iopub.status.idle": "2020-10-22T19:24:25.553584Z",
     "shell.execute_reply": "2020-10-22T19:24:25.554166Z"
    },
    "papermill": {
     "duration": 0.156932,
     "end_time": "2020-10-22T19:24:25.554313",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.397381",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2400, 68, 5)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_y = np.array(train_df[pred_cols].values.tolist()).transpose((0, 2, 1))\n",
    "train_y.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.033152,
     "end_time": "2020-10-22T19:24:25.619968",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.586816",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Sequence Features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.03394,
     "end_time": "2020-10-22T19:24:25.685995",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.652055",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Each mRNA contains 107 bases in train:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:25.769039Z",
     "iopub.status.busy": "2020-10-22T19:24:25.768049Z",
     "iopub.status.idle": "2020-10-22T19:24:25.772551Z",
     "shell.execute_reply": "2020-10-22T19:24:25.773074Z"
    },
    "papermill": {
     "duration": 0.052819,
     "end_time": "2020-10-22T19:24:25.773228",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.720409",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>sequence</th>\n",
       "      <th>structure</th>\n",
       "      <th>predicted_loop_type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>id_001f94081</td>\n",
       "      <td>GGAAAAGCUCUAAUAACAGGAGACUAGGACUACGUAUUUCUAGGUA...</td>\n",
       "      <td>.....((((((.......)))).)).((.....((..((((((......</td>\n",
       "      <td>EEEEESSSSSSHHHHHHHSSSSBSSXSSIIIIISSIISSSSSSHHH...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>id_0049f53ba</td>\n",
       "      <td>GGAAAAAGCGCGCGCGGUUAGCGCGCGCUUUUGCGCGCGCUGUACC...</td>\n",
       "      <td>.....(((((((((((((((((((((((....)))))))))).)))...</td>\n",
       "      <td>EEEEESSSSSSSSSSSSSSSSSSSSSSSHHHHSSSSSSSSSSBSSS...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>id_006f36f57</td>\n",
       "      <td>GGAAAGUGCUCAGAUAAGCUAAGCUCGAAUAGCAAUCGAAUAGAAU...</td>\n",
       "      <td>.....((((.((.....((((.(((.....)))..((((......)...</td>\n",
       "      <td>EEEEESSSSISSIIIIISSSSMSSSHHHHHSSSMMSSSSHHHHHHS...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>id_0082d463b</td>\n",
       "      <td>GGAAAAGCGCGCGCGCGCGCGCGAAAAAGCGCGCGCGCGCGCGCGC...</td>\n",
       "      <td>......((((((((((((((((......))))))))))))))))((...</td>\n",
       "      <td>EEEEEESSSSSSSSSSSSSSSSHHHHHHSSSSSSSSSSSSSSSSSS...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>id_0087940f4</td>\n",
       "      <td>GGAAAAUAUAUAAUAUAUUAUAUAAAUAUAUUAUAGAAGUAUAAUA...</td>\n",
       "      <td>.....(((((((.((((((((((((.(((((((((....)))))))...</td>\n",
       "      <td>EEEEESSSSSSSBSSSSSSSSSSSSBSSSSSSSSSHHHHSSSSSSS...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             id                                           sequence  \\\n",
       "0  id_001f94081  GGAAAAGCUCUAAUAACAGGAGACUAGGACUACGUAUUUCUAGGUA...   \n",
       "1  id_0049f53ba  GGAAAAAGCGCGCGCGGUUAGCGCGCGCUUUUGCGCGCGCUGUACC...   \n",
       "2  id_006f36f57  GGAAAGUGCUCAGAUAAGCUAAGCUCGAAUAGCAAUCGAAUAGAAU...   \n",
       "3  id_0082d463b  GGAAAAGCGCGCGCGCGCGCGCGAAAAAGCGCGCGCGCGCGCGCGC...   \n",
       "4  id_0087940f4  GGAAAAUAUAUAAUAUAUUAUAUAAAUAUAUUAUAGAAGUAUAAUA...   \n",
       "\n",
       "                                           structure  \\\n",
       "0  .....((((((.......)))).)).((.....((..((((((......   \n",
       "1  .....(((((((((((((((((((((((....)))))))))).)))...   \n",
       "2  .....((((.((.....((((.(((.....)))..((((......)...   \n",
       "3  ......((((((((((((((((......))))))))))))))))((...   \n",
       "4  .....(((((((.((((((((((((.(((((((((....)))))))...   \n",
       "\n",
       "                                 predicted_loop_type  \n",
       "0  EEEEESSSSSSHHHHHHHSSSSBSSXSSIIIIISSIISSSSSSHHH...  \n",
       "1  EEEEESSSSSSSSSSSSSSSSSSSSSSSHHHHSSSSSSSSSSBSSS...  \n",
       "2  EEEEESSSSISSIIIIISSSSMSSSHHHHHSSSMMSSSSHHHHHHS...  \n",
       "3  EEEEEESSSSSSSSSSSSSSSSHHHHHHSSSSSSSSSSSSSSSSSS...  \n",
       "4  EEEEESSSSSSSBSSSSSSSSSSSSBSSSSSSSSSHHHHSSSSSSS...  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df[[\"id\", \"sequence\", \"structure\", \"predicted_loop_type\"]].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.034427,
     "end_time": "2020-10-22T19:24:25.842196",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.807769",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Both `sequence`, `structure` and `predicted_loop_type` are sequences data for those 107 bases, let's transform them. \n",
    "\n",
    "(2400, 3) -> (2400, 107, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:25.926120Z",
     "iopub.status.busy": "2020-10-22T19:24:25.925058Z",
     "iopub.status.idle": "2020-10-22T19:24:26.832121Z",
     "shell.execute_reply": "2020-10-22T19:24:26.832704Z"
    },
    "papermill": {
     "duration": 0.955207,
     "end_time": "2020-10-22T19:24:26.832867",
     "exception": false,
     "start_time": "2020-10-22T19:24:25.877660",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((2400, 107, 3), (629, 107, 3), (3005, 130, 3))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# label encodings for \n",
    "token2int = {x:i for i, x in enumerate('().ACGUBEHIMSX')}\n",
    "sequence_token2int = {x:i for i, x in enumerate('AGUC')}\n",
    "structure_token2int = {\n",
    "    '.': 0,\n",
    "    '(': 1,\n",
    "    ')': 2,\n",
    "}\n",
    "loop_token2int = {x:i for i, x in enumerate('SMIBHEX')}\n",
    "token2int_map = {\n",
    "    \"sequence\": sequence_token2int,\n",
    "    \"structure\": structure_token2int,\n",
    "    \"predicted_loop_type\": loop_token2int\n",
    "}\n",
    "sequence_columns = [\"sequence\", \"structure\", \"predicted_loop_type\"]\n",
    "\n",
    "def to_seq(df):\n",
    "    return np.transpose(\n",
    "        np.array(\n",
    "            df[sequence_columns]\n",
    "            .applymap(lambda seq: [token2int[x] for x in seq])\n",
    "            .values\n",
    "            .tolist()\n",
    "        ),\n",
    "        (0, 2, 1)\n",
    "    )\n",
    "\n",
    "train = to_seq(train_df)\n",
    "public = to_seq(public_df)\n",
    "private = to_seq(private_df)\n",
    "\n",
    "train.shape, public.shape, private.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.035225,
     "end_time": "2020-10-22T19:24:26.903796",
     "exception": false,
     "start_time": "2020-10-22T19:24:26.868571",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Now let's apply one-hot encoding to the data\n",
    "\n",
    "(2400, 107, 3) -> (2400, 107, 14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:27.033251Z",
     "iopub.status.busy": "2020-10-22T19:24:27.004166Z",
     "iopub.status.idle": "2020-10-22T19:24:27.845754Z",
     "shell.execute_reply": "2020-10-22T19:24:27.846331Z"
    },
    "papermill": {
     "duration": 0.909446,
     "end_time": "2020-10-22T19:24:27.846481",
     "exception": false,
     "start_time": "2020-10-22T19:24:26.937035",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((2400, 107, 14), (629, 107, 14), (3005, 130, 14))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def to_one_hot(df):\n",
    "    temp = np.transpose(\n",
    "        np.array([\n",
    "            df[col]\n",
    "            .apply(lambda seq: [token2int_map[col][x] for x in seq])\n",
    "            .values\n",
    "            .tolist()\n",
    "            for col in sequence_columns\n",
    "        ]),\n",
    "        (1, 2, 0)\n",
    "    )\n",
    "    ohe_1 = tf.keras.utils.to_categorical(temp[:,:,0], 4)\n",
    "    ohe_2 = tf.keras.utils.to_categorical(temp[:,:,1], 3)\n",
    "    ohe_3 = tf.keras.utils.to_categorical(temp[:,:,2], 7)\n",
    "    return np.concatenate([ohe_1, ohe_2, ohe_3], axis=2)\n",
    "\n",
    "train_ohe = to_one_hot(train_df)\n",
    "public_ohe = to_one_hot(public_df)\n",
    "private_ohe = to_one_hot(private_df)\n",
    "\n",
    "train_ohe.shape, public_ohe.shape, private_ohe.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.033904,
     "end_time": "2020-10-22T19:24:27.916855",
     "exception": false,
     "start_time": "2020-10-22T19:24:27.882951",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Pair/Connectivity Features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.034858,
     "end_time": "2020-10-22T19:24:27.986715",
     "exception": false,
     "start_time": "2020-10-22T19:24:27.951857",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Get the adjacency matrices from the structure sequence:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:28.073401Z",
     "iopub.status.busy": "2020-10-22T19:24:28.066056Z",
     "iopub.status.idle": "2020-10-22T19:24:30.335569Z",
     "shell.execute_reply": "2020-10-22T19:24:30.336133Z"
    },
    "papermill": {
     "duration": 2.31473,
     "end_time": "2020-10-22T19:24:30.336281",
     "exception": false,
     "start_time": "2020-10-22T19:24:28.021551",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((2400, 107, 107), (629, 107, 107), (3005, 130, 130))"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_adjacency_matrix(inps):\n",
    "    As = []\n",
    "    for row in range(0, inps.shape[0]):\n",
    "        A = np.zeros((inps.shape[1], inps.shape[1]))\n",
    "        stack = []\n",
    "        opened_so_far = []\n",
    "\n",
    "        for seqpos in range(0, inps.shape[1]):\n",
    "            # A[seqpos, seqpos] = 1\n",
    "            if inps[row, seqpos, 1] == 0:\n",
    "                stack.append(seqpos)\n",
    "                opened_so_far.append(seqpos)\n",
    "            elif inps[row, seqpos, 1] == 1:\n",
    "                openpos = stack.pop()\n",
    "                A[openpos, seqpos] = 1\n",
    "                A[seqpos, openpos] = 1\n",
    "        As.append(A)\n",
    "    return np.array(As)\n",
    "\n",
    "train_adj = get_adjacency_matrix(train)\n",
    "public_adj = get_adjacency_matrix(public)\n",
    "private_adj = get_adjacency_matrix(private)\n",
    "\n",
    "train_adj.shape, public_adj.shape, private_adj.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:30.413716Z",
     "iopub.status.busy": "2020-10-22T19:24:30.412498Z",
     "iopub.status.idle": "2020-10-22T19:24:30.477546Z",
     "shell.execute_reply": "2020-10-22T19:24:30.478175Z"
    },
    "papermill": {
     "duration": 0.107886,
     "end_time": "2020-10-22T19:24:30.478326",
     "exception": false,
     "start_time": "2020-10-22T19:24:30.370440",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0041319474772178065, 0.004322480243829655, 0.0038859494530811564)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_adj.mean(), public_adj.mean(), private_adj.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.038569,
     "end_time": "2020-10-22T19:24:30.557278",
     "exception": false,
     "start_time": "2020-10-22T19:24:30.518709",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Get the base pair probabilities from the given bpps files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:30.638093Z",
     "iopub.status.busy": "2020-10-22T19:24:30.637090Z",
     "iopub.status.idle": "2020-10-22T19:24:49.122285Z",
     "shell.execute_reply": "2020-10-22T19:24:49.121237Z"
    },
    "papermill": {
     "duration": 18.529021,
     "end_time": "2020-10-22T19:24:49.122421",
     "exception": false,
     "start_time": "2020-10-22T19:24:30.593400",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2400/2400 [00:07<00:00, 335.65it/s]\n",
      "100%|██████████| 629/629 [00:01<00:00, 340.73it/s]\n",
      "100%|██████████| 3005/3005 [00:09<00:00, 326.04it/s]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((2400, 107, 107), (629, 107, 107), (3005, 130, 130))"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_bpps(mRNA_ids):\n",
    "    bpps = []\n",
    "    for mRNA_id in tqdm(mRNA_ids):\n",
    "        bpps.append(\n",
    "            np.load(f\"{bpps_path}/{mRNA_id}.npy\"),\n",
    "        )\n",
    "    return np.array(bpps)\n",
    "\n",
    "\n",
    "train_bpps = get_bpps(train_df.id.values)\n",
    "public_bpps = get_bpps(public_df.id.values)\n",
    "private_bpps = get_bpps(private_df.id.values)\n",
    "\n",
    "train_bpps.shape, public_bpps.shape, private_bpps.shape "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:49.346532Z",
     "iopub.status.busy": "2020-10-22T19:24:49.345319Z",
     "iopub.status.idle": "2020-10-22T19:24:49.411792Z",
     "shell.execute_reply": "2020-10-22T19:24:49.412407Z"
    },
    "papermill": {
     "duration": 0.182027,
     "end_time": "2020-10-22T19:24:49.412576",
     "exception": false,
     "start_time": "2020-10-22T19:24:49.230549",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.003936093904915271, 0.0037162253979562986, 0.004133111640768379)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_bpps.mean(), public_bpps.mean(), private_bpps.mean() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.10521,
     "end_time": "2020-10-22T19:24:49.622297",
     "exception": false,
     "start_time": "2020-10-22T19:24:49.517087",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Get pair probabilities statistics for each base"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:49.838141Z",
     "iopub.status.busy": "2020-10-22T19:24:49.836892Z",
     "iopub.status.idle": "2020-10-22T19:24:50.026842Z",
     "shell.execute_reply": "2020-10-22T19:24:50.026229Z"
    },
    "papermill": {
     "duration": 0.300109,
     "end_time": "2020-10-22T19:24:50.026956",
     "exception": false,
     "start_time": "2020-10-22T19:24:49.726847",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_bpps_stats = [train_bpps.mean(axis=2), train_bpps.max(axis=2)]\n",
    "public_bpps_stats = [public_bpps.mean(axis=2), public_bpps.max(axis=2)]\n",
    "private_bpps_stats = [private_bpps.mean(axis=2), private_bpps.max(axis=2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:50.240370Z",
     "iopub.status.busy": "2020-10-22T19:24:50.239232Z",
     "iopub.status.idle": "2020-10-22T19:24:50.251079Z",
     "shell.execute_reply": "2020-10-22T19:24:50.250515Z"
    },
    "papermill": {
     "duration": 0.121859,
     "end_time": "2020-10-22T19:24:50.251189",
     "exception": false,
     "start_time": "2020-10-22T19:24:50.129330",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((2400, 107, 2), (629, 107, 2), (3005, 130, 2))"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_bpps_stats = np.concatenate([stats[:,:,None] for stats in train_bpps_stats], axis=2)\n",
    "public_bpps_stats = np.concatenate([stats[:,:,None] for stats in public_bpps_stats], axis=2)\n",
    "private_bpps_stats = np.concatenate([stats[:,:,None] for stats in private_bpps_stats], axis=2)\n",
    "\n",
    "train_bpps_stats.shape, public_bpps_stats.shape, private_bpps_stats.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.104033,
     "end_time": "2020-10-22T19:24:50.462873",
     "exception": false,
     "start_time": "2020-10-22T19:24:50.358840",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.107355,
     "end_time": "2020-10-22T19:24:50.679824",
     "exception": false,
     "start_time": "2020-10-22T19:24:50.572469",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F690886%2F83da03ea10245b6867febd170f6188f4%2FScreenshot%202020-10-08%20at%207.51.09%20AM.png?generation=1602114976893373&alt=media)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:50.919935Z",
     "iopub.status.busy": "2020-10-22T19:24:50.918976Z",
     "iopub.status.idle": "2020-10-22T19:24:56.926498Z",
     "shell.execute_reply": "2020-10-22T19:24:56.927058Z"
    },
    "papermill": {
     "duration": 6.140157,
     "end_time": "2020-10-22T19:24:56.927204",
     "exception": false,
     "start_time": "2020-10-22T19:24:50.787047",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scored_seq_length = 68\n",
    "\n",
    "# loss functions\n",
    "def rmse(y_actual, y_pred):\n",
    "    mse = tf.keras.losses.mean_squared_error(y_actual, y_pred)\n",
    "    return K.sqrt(mse)\n",
    "\n",
    "def mcrmse(y_actual, y_pred):\n",
    "    score = 0\n",
    "    for i in range(y_actual.shape[2]):\n",
    "        score += rmse(y_actual[:, :, i], y_pred[:, :, i]) / y_actual.shape[2]\n",
    "    return score\n",
    "    \n",
    "\n",
    "def build_model(input_seq_len=107, output_seq_len=scored_seq_length):\n",
    "    \n",
    "    def _bi_gru_block(x, hidden_dim, dropout):\n",
    "        gru = L.Bidirectional(\n",
    "            L.GRU(hidden_dim, \n",
    "                  dropout=dropout,\n",
    "                  return_sequences=True,\n",
    "                 ),\n",
    "        )(x)\n",
    "        return gru\n",
    "\n",
    "    def _conv_block(x, adj_m, bpp_m, conv_filters, graph_channels):\n",
    "        # local 1-D convolution\n",
    "        conv = L.Conv1D(\n",
    "            conv_filters, 5,\n",
    "            padding='same',\n",
    "            activation='tanh',\n",
    "        )(x)\n",
    "        \n",
    "        # graph convolution\n",
    "        gcn_1 = GraphConv(\n",
    "            graph_channels,\n",
    "            activation='tanh',\n",
    "        )([conv, adj_m])\n",
    "        \n",
    "        gcn_2 = GraphConv(\n",
    "            graph_channels,\n",
    "            activation='tanh',\n",
    "        )([conv, bpp_m])\n",
    "\n",
    "        conv = L.Concatenate()([conv, gcn_1, gcn_2])\n",
    "        conv = L.Activation(\"relu\")(conv)\n",
    "        conv = L.SpatialDropout1D(0.1)(conv)\n",
    "        \n",
    "        return conv\n",
    "    \n",
    "    # inputs\n",
    "    one_hot_encoding_inputs = L.Input(shape=(input_seq_len, 14), name=\"onehot\")\n",
    "    # adjacency matrix about seq. connectivity\n",
    "    adj_matrix_inputs = L.Input((input_seq_len, input_seq_len), name=\"adjmatrix\")\n",
    "    # base pair proba\n",
    "    base_pair_proba_inputs = L.Input((input_seq_len, input_seq_len), name=\"pairproba\")\n",
    "    # base pair proba stats\n",
    "    base_pair_proba_stats_inputs = L.Input(shape=(input_seq_len, 2), name=\"pairprobastats\")\n",
    "    \n",
    "    merged_inputs = L.Concatenate()([one_hot_encoding_inputs, base_pair_proba_stats_inputs])\n",
    "    \n",
    "    # convolution and recurrent blocks.\n",
    "    hidden = _conv_block(merged_inputs, adj_matrix_inputs, base_pair_proba_inputs, 512, 80)\n",
    "    hidden = _bi_gru_block(hidden, 256, 0.5)\n",
    "    hidden = _conv_block(hidden, adj_matrix_inputs, base_pair_proba_inputs, 512, 80)\n",
    "    hidden = _bi_gru_block(hidden, 256, 0.5)\n",
    "    \n",
    "    out = hidden[:, :output_seq_len]\n",
    "    out = L.Dense(5, activation='linear')(out)\n",
    "    \n",
    "    model = tf.keras.Model(\n",
    "        inputs=[\n",
    "            one_hot_encoding_inputs,\n",
    "            adj_matrix_inputs,\n",
    "            base_pair_proba_inputs,\n",
    "            base_pair_proba_stats_inputs,\n",
    "        ],\n",
    "        outputs=out,\n",
    "    )\n",
    "\n",
    "    return model\n",
    "\n",
    "model = build_model()\n",
    "plot_model(model, to_file='model_plot.png', show_shapes=True, show_layer_names=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.114057,
     "end_time": "2020-10-22T19:24:57.150902",
     "exception": false,
     "start_time": "2020-10-22T19:24:57.036845",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:57.388485Z",
     "iopub.status.busy": "2020-10-22T19:24:57.386945Z",
     "iopub.status.idle": "2020-10-22T19:24:57.540648Z",
     "shell.execute_reply": "2020-10-22T19:24:57.541398Z"
    },
    "papermill": {
     "duration": 0.275731,
     "end_time": "2020-10-22T19:24:57.541560",
     "exception": false,
     "start_time": "2020-10-22T19:24:57.265829",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(2160, 107, 14),\n",
       " (240, 107, 14),\n",
       " (2160, 107, 107),\n",
       " (240, 107, 107),\n",
       " (2160, 107, 107),\n",
       " (240, 107, 107),\n",
       " (2160, 107, 2),\n",
       " (240, 107, 2),\n",
       " (2160, 68, 5),\n",
       " (240, 68, 5),\n",
       " (2160,),\n",
       " (240,),\n",
       " (2160,),\n",
       " (240,)]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "split_results  = train_test_split(\n",
    "    train_ohe,\n",
    "    train_adj,\n",
    "    train_bpps,\n",
    "    train_bpps_stats,\n",
    "    train_y,\n",
    "    train_df.signal_to_noise,\n",
    "    train_df.SN_filter,\n",
    "    test_size=0.1,\n",
    "    random_state=42,\n",
    ")\n",
    "\n",
    "[a.shape for a in split_results]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:57.806914Z",
     "iopub.status.busy": "2020-10-22T19:24:57.805492Z",
     "iopub.status.idle": "2020-10-22T19:24:57.807721Z",
     "shell.execute_reply": "2020-10-22T19:24:57.808286Z"
    },
    "papermill": {
     "duration": 0.152446,
     "end_time": "2020-10-22T19:24:57.808434",
     "exception": false,
     "start_time": "2020-10-22T19:24:57.655988",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "trn_ohe, val_ohe, trn_adj, val_adj, trn_bpps, val_bpps, trn_bpps_stats, val_bpps_stats, trn_y, val_y, trn_snr, val_snr, trn_snf, val_snf = split_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:58.057324Z",
     "iopub.status.busy": "2020-10-22T19:24:58.055475Z",
     "iopub.status.idle": "2020-10-22T19:24:59.498407Z",
     "shell.execute_reply": "2020-10-22T19:24:59.497437Z"
    },
    "papermill": {
     "duration": 1.572406,
     "end_time": "2020-10-22T19:24:59.498532",
     "exception": false,
     "start_time": "2020-10-22T19:24:57.926126",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model = build_model()\n",
    "model.compile(tf.keras.optimizers.Adam(), loss=mcrmse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:59.732382Z",
     "iopub.status.busy": "2020-10-22T19:24:59.731377Z",
     "iopub.status.idle": "2020-10-22T19:24:59.736190Z",
     "shell.execute_reply": "2020-10-22T19:24:59.735558Z"
    },
    "papermill": {
     "duration": 0.125857,
     "end_time": "2020-10-22T19:24:59.736301",
     "exception": false,
     "start_time": "2020-10-22T19:24:59.610444",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "trn_inputs = [trn_ohe, trn_adj, trn_bpps, trn_bpps_stats]\n",
    "val_inputs = [val_ohe, val_adj, val_bpps, val_bpps_stats]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:24:59.974320Z",
     "iopub.status.busy": "2020-10-22T19:24:59.972932Z",
     "iopub.status.idle": "2020-10-22T19:24:59.989757Z",
     "shell.execute_reply": "2020-10-22T19:24:59.989103Z"
    },
    "papermill": {
     "duration": 0.138136,
     "end_time": "2020-10-22T19:24:59.989903",
     "exception": false,
     "start_time": "2020-10-22T19:24:59.851767",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# only validate on data with sn_filter = 1\n",
    "val_mask = np.where((val_snf==1))\n",
    "val_inputs = [val_input[val_mask] for val_input in val_inputs]\n",
    "val_y = val_y[val_mask]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:25:00.224042Z",
     "iopub.status.busy": "2020-10-22T19:25:00.223218Z",
     "iopub.status.idle": "2020-10-22T19:25:00.228012Z",
     "shell.execute_reply": "2020-10-22T19:25:00.227397Z"
    },
    "papermill": {
     "duration": 0.124407,
     "end_time": "2020-10-22T19:25:00.228132",
     "exception": false,
     "start_time": "2020-10-22T19:25:00.103725",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "sample_weight = np.log(trn_snr+1.11)/2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:25:00.470772Z",
     "iopub.status.busy": "2020-10-22T19:25:00.469747Z",
     "iopub.status.idle": "2020-10-22T19:31:49.062965Z",
     "shell.execute_reply": "2020-10-22T19:31:49.063510Z"
    },
    "papermill": {
     "duration": 408.716209,
     "end_time": "2020-10-22T19:31:49.063726",
     "exception": false,
     "start_time": "2020-10-22T19:25:00.347517",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/300\n",
      "34/34 - 5s - loss: 0.3642 - val_loss: 0.3536\n",
      "Epoch 2/300\n",
      "34/34 - 3s - loss: 0.2819 - val_loss: 0.3102\n",
      "Epoch 3/300\n",
      "34/34 - 3s - loss: 0.2584 - val_loss: 0.2873\n",
      "Epoch 4/300\n",
      "34/34 - 3s - loss: 0.2468 - val_loss: 0.2689\n",
      "Epoch 5/300\n",
      "34/34 - 3s - loss: 0.2388 - val_loss: 0.2581\n",
      "Epoch 6/300\n",
      "34/34 - 3s - loss: 0.2319 - val_loss: 0.2500\n",
      "Epoch 7/300\n",
      "34/34 - 3s - loss: 0.2255 - val_loss: 0.2458\n",
      "Epoch 8/300\n",
      "34/34 - 3s - loss: 0.2210 - val_loss: 0.2435\n",
      "Epoch 9/300\n",
      "34/34 - 3s - loss: 0.2170 - val_loss: 0.2414\n",
      "Epoch 10/300\n",
      "34/34 - 3s - loss: 0.2147 - val_loss: 0.2392\n",
      "Epoch 11/300\n",
      "34/34 - 3s - loss: 0.2106 - val_loss: 0.2391\n",
      "Epoch 12/300\n",
      "34/34 - 3s - loss: 0.2086 - val_loss: 0.2340\n",
      "Epoch 13/300\n",
      "34/34 - 3s - loss: 0.2056 - val_loss: 0.2326\n",
      "Epoch 14/300\n",
      "34/34 - 3s - loss: 0.2033 - val_loss: 0.2308\n",
      "Epoch 15/300\n",
      "34/34 - 3s - loss: 0.2030 - val_loss: 0.2306\n",
      "Epoch 16/300\n",
      "34/34 - 3s - loss: 0.2005 - val_loss: 0.2282\n",
      "Epoch 17/300\n",
      "34/34 - 3s - loss: 0.1978 - val_loss: 0.2265\n",
      "Epoch 18/300\n",
      "34/34 - 3s - loss: 0.1969 - val_loss: 0.2248\n",
      "Epoch 19/300\n",
      "34/34 - 2s - loss: 0.1957 - val_loss: 0.2260\n",
      "Epoch 20/300\n",
      "34/34 - 3s - loss: 0.1941 - val_loss: 0.2225\n",
      "Epoch 21/300\n",
      "34/34 - 2s - loss: 0.1924 - val_loss: 0.2227\n",
      "Epoch 22/300\n",
      "34/34 - 3s - loss: 0.1902 - val_loss: 0.2178\n",
      "Epoch 23/300\n",
      "34/34 - 2s - loss: 0.1892 - val_loss: 0.2206\n",
      "Epoch 24/300\n",
      "34/34 - 3s - loss: 0.1881 - val_loss: 0.2187\n",
      "Epoch 25/300\n",
      "34/34 - 3s - loss: 0.1862 - val_loss: 0.2182\n",
      "Epoch 26/300\n",
      "34/34 - 3s - loss: 0.1853 - val_loss: 0.2175\n",
      "Epoch 27/300\n",
      "34/34 - 3s - loss: 0.1847 - val_loss: 0.2136\n",
      "Epoch 28/300\n",
      "34/34 - 3s - loss: 0.1845 - val_loss: 0.2179\n",
      "Epoch 29/300\n",
      "34/34 - 3s - loss: 0.1810 - val_loss: 0.2184\n",
      "Epoch 30/300\n",
      "34/34 - 3s - loss: 0.1811 - val_loss: 0.2134\n",
      "Epoch 31/300\n",
      "34/34 - 3s - loss: 0.1796 - val_loss: 0.2114\n",
      "Epoch 32/300\n",
      "34/34 - 2s - loss: 0.1782 - val_loss: 0.2122\n",
      "Epoch 33/300\n",
      "34/34 - 3s - loss: 0.1778 - val_loss: 0.2116\n",
      "Epoch 34/300\n",
      "34/34 - 2s - loss: 0.1765 - val_loss: 0.2128\n",
      "Epoch 35/300\n",
      "34/34 - 3s - loss: 0.1757 - val_loss: 0.2107\n",
      "Epoch 36/300\n",
      "34/34 - 3s - loss: 0.1751 - val_loss: 0.2089\n",
      "Epoch 37/300\n",
      "34/34 - 3s - loss: 0.1744 - val_loss: 0.2118\n",
      "Epoch 38/300\n",
      "34/34 - 2s - loss: 0.1729 - val_loss: 0.2103\n",
      "Epoch 39/300\n",
      "34/34 - 3s - loss: 0.1719 - val_loss: 0.2085\n",
      "Epoch 40/300\n",
      "34/34 - 3s - loss: 0.1722 - val_loss: 0.2069\n",
      "Epoch 41/300\n",
      "34/34 - 3s - loss: 0.1707 - val_loss: 0.2061\n",
      "Epoch 42/300\n",
      "34/34 - 2s - loss: 0.1695 - val_loss: 0.2068\n",
      "Epoch 43/300\n",
      "34/34 - 2s - loss: 0.1694 - val_loss: 0.2070\n",
      "Epoch 44/300\n",
      "34/34 - 3s - loss: 0.1682 - val_loss: 0.2056\n",
      "Epoch 45/300\n",
      "34/34 - 3s - loss: 0.1686 - val_loss: 0.2077\n",
      "Epoch 46/300\n",
      "34/34 - 3s - loss: 0.1668 - val_loss: 0.2050\n",
      "Epoch 47/300\n",
      "34/34 - 3s - loss: 0.1656 - val_loss: 0.2046\n",
      "Epoch 48/300\n",
      "34/34 - 3s - loss: 0.1652 - val_loss: 0.2034\n",
      "Epoch 49/300\n",
      "34/34 - 3s - loss: 0.1642 - val_loss: 0.2043\n",
      "Epoch 50/300\n",
      "34/34 - 3s - loss: 0.1646 - val_loss: 0.2063\n",
      "Epoch 51/300\n",
      "34/34 - 2s - loss: 0.1625 - val_loss: 0.2055\n",
      "Epoch 52/300\n",
      "34/34 - 3s - loss: 0.1619 - val_loss: 0.2020\n",
      "Epoch 53/300\n",
      "34/34 - 3s - loss: 0.1610 - val_loss: 0.2012\n",
      "Epoch 54/300\n",
      "34/34 - 3s - loss: 0.1607 - val_loss: 0.2022\n",
      "Epoch 55/300\n",
      "34/34 - 2s - loss: 0.1599 - val_loss: 0.2025\n",
      "Epoch 56/300\n",
      "34/34 - 3s - loss: 0.1595 - val_loss: 0.2003\n",
      "Epoch 57/300\n",
      "34/34 - 2s - loss: 0.1585 - val_loss: 0.2031\n",
      "Epoch 58/300\n",
      "34/34 - 3s - loss: 0.1585 - val_loss: 0.2015\n",
      "Epoch 59/300\n",
      "34/34 - 2s - loss: 0.1584 - val_loss: 0.2010\n",
      "Epoch 60/300\n",
      "34/34 - 2s - loss: 0.1574 - val_loss: 0.2009\n",
      "Epoch 61/300\n",
      "34/34 - 3s - loss: 0.1560 - val_loss: 0.2001\n",
      "Epoch 62/300\n",
      "34/34 - 3s - loss: 0.1558 - val_loss: 0.2018\n",
      "Epoch 63/300\n",
      "34/34 - 2s - loss: 0.1555 - val_loss: 0.2028\n",
      "Epoch 64/300\n",
      "34/34 - 3s - loss: 0.1550 - val_loss: 0.1995\n",
      "Epoch 65/300\n",
      "34/34 - 2s - loss: 0.1539 - val_loss: 0.2005\n",
      "Epoch 66/300\n",
      "34/34 - 3s - loss: 0.1532 - val_loss: 0.1982\n",
      "Epoch 67/300\n",
      "34/34 - 2s - loss: 0.1528 - val_loss: 0.2010\n",
      "Epoch 68/300\n",
      "34/34 - 2s - loss: 0.1520 - val_loss: 0.1989\n",
      "Epoch 69/300\n",
      "34/34 - 2s - loss: 0.1524 - val_loss: 0.1987\n",
      "Epoch 70/300\n",
      "34/34 - 3s - loss: 0.1510 - val_loss: 0.2010\n",
      "Epoch 71/300\n",
      "34/34 - 2s - loss: 0.1518 - val_loss: 0.2030\n",
      "Epoch 72/300\n",
      "34/34 - 3s - loss: 0.1511 - val_loss: 0.1978\n",
      "Epoch 73/300\n",
      "34/34 - 3s - loss: 0.1500 - val_loss: 0.2013\n",
      "Epoch 74/300\n",
      "34/34 - 3s - loss: 0.1493 - val_loss: 0.1971\n",
      "Epoch 75/300\n",
      "34/34 - 2s - loss: 0.1483 - val_loss: 0.1981\n",
      "Epoch 76/300\n",
      "34/34 - 2s - loss: 0.1482 - val_loss: 0.1984\n",
      "Epoch 77/300\n",
      "34/34 - 2s - loss: 0.1479 - val_loss: 0.1992\n",
      "Epoch 78/300\n",
      "34/34 - 3s - loss: 0.1475 - val_loss: 0.2007\n",
      "Epoch 79/300\n",
      "34/34 - 3s - loss: 0.1461 - val_loss: 0.1957\n",
      "Epoch 80/300\n",
      "34/34 - 2s - loss: 0.1460 - val_loss: 0.2019\n",
      "Epoch 81/300\n",
      "34/34 - 2s - loss: 0.1456 - val_loss: 0.1981\n",
      "Epoch 82/300\n",
      "34/34 - 2s - loss: 0.1451 - val_loss: 0.1973\n",
      "Epoch 83/300\n",
      "34/34 - 3s - loss: 0.1455 - val_loss: 0.1987\n",
      "Epoch 84/300\n",
      "34/34 - 2s - loss: 0.1443 - val_loss: 0.1968\n",
      "Epoch 85/300\n",
      "34/34 - 3s - loss: 0.1436 - val_loss: 0.1972\n",
      "Epoch 86/300\n",
      "34/34 - 2s - loss: 0.1431 - val_loss: 0.1989\n",
      "Epoch 87/300\n",
      "34/34 - 2s - loss: 0.1435 - val_loss: 0.1968\n",
      "Epoch 88/300\n",
      "34/34 - 2s - loss: 0.1425 - val_loss: 0.1969\n",
      "Epoch 89/300\n",
      "\n",
      "Epoch 00089: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n",
      "34/34 - 2s - loss: 0.1427 - val_loss: 0.1970\n",
      "Epoch 90/300\n",
      "34/34 - 3s - loss: 0.1392 - val_loss: 0.1954\n",
      "Epoch 91/300\n",
      "34/34 - 3s - loss: 0.1377 - val_loss: 0.1942\n",
      "Epoch 92/300\n",
      "34/34 - 3s - loss: 0.1369 - val_loss: 0.1940\n",
      "Epoch 93/300\n",
      "34/34 - 2s - loss: 0.1367 - val_loss: 0.1940\n",
      "Epoch 94/300\n",
      "34/34 - 3s - loss: 0.1362 - val_loss: 0.1938\n",
      "Epoch 95/300\n",
      "34/34 - 3s - loss: 0.1362 - val_loss: 0.1939\n",
      "Epoch 96/300\n",
      "34/34 - 3s - loss: 0.1358 - val_loss: 0.1938\n",
      "Epoch 97/300\n",
      "34/34 - 3s - loss: 0.1356 - val_loss: 0.1938\n",
      "Epoch 98/300\n",
      "34/34 - 2s - loss: 0.1354 - val_loss: 0.1947\n",
      "Epoch 99/300\n",
      "34/34 - 3s - loss: 0.1352 - val_loss: 0.1943\n",
      "Epoch 100/300\n",
      "34/34 - 2s - loss: 0.1350 - val_loss: 0.1940\n",
      "Epoch 101/300\n",
      "34/34 - 3s - loss: 0.1349 - val_loss: 0.1934\n",
      "Epoch 102/300\n",
      "34/34 - 2s - loss: 0.1348 - val_loss: 0.1936\n",
      "Epoch 103/300\n",
      "34/34 - 3s - loss: 0.1345 - val_loss: 0.1935\n",
      "Epoch 104/300\n",
      "34/34 - 3s - loss: 0.1344 - val_loss: 0.1936\n",
      "Epoch 105/300\n",
      "34/34 - 2s - loss: 0.1341 - val_loss: 0.1936\n",
      "Epoch 106/300\n",
      "34/34 - 3s - loss: 0.1342 - val_loss: 0.1932\n",
      "Epoch 107/300\n",
      "34/34 - 2s - loss: 0.1343 - val_loss: 0.1934\n",
      "Epoch 108/300\n",
      "34/34 - 3s - loss: 0.1340 - val_loss: 0.1932\n",
      "Epoch 109/300\n",
      "34/34 - 3s - loss: 0.1339 - val_loss: 0.1936\n",
      "Epoch 110/300\n",
      "34/34 - 3s - loss: 0.1339 - val_loss: 0.1931\n",
      "Epoch 111/300\n",
      "34/34 - 2s - loss: 0.1336 - val_loss: 0.1937\n",
      "Epoch 112/300\n",
      "34/34 - 3s - loss: 0.1336 - val_loss: 0.1932\n",
      "Epoch 113/300\n",
      "34/34 - 3s - loss: 0.1338 - val_loss: 0.1931\n",
      "Epoch 114/300\n",
      "34/34 - 2s - loss: 0.1336 - val_loss: 0.1935\n",
      "Epoch 115/300\n",
      "34/34 - 2s - loss: 0.1333 - val_loss: 0.1935\n",
      "Epoch 116/300\n",
      "\n",
      "Epoch 00116: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n",
      "34/34 - 3s - loss: 0.1331 - val_loss: 0.1931\n",
      "Epoch 117/300\n",
      "34/34 - 2s - loss: 0.1327 - val_loss: 0.1932\n",
      "Epoch 118/300\n",
      "34/34 - 3s - loss: 0.1327 - val_loss: 0.1932\n",
      "Epoch 119/300\n",
      "34/34 - 2s - loss: 0.1325 - val_loss: 0.1931\n",
      "Epoch 120/300\n",
      "34/34 - 3s - loss: 0.1327 - val_loss: 0.1930\n",
      "Epoch 121/300\n",
      "34/34 - 3s - loss: 0.1330 - val_loss: 0.1932\n",
      "Epoch 122/300\n",
      "34/34 - 3s - loss: 0.1329 - val_loss: 0.1932\n",
      "Epoch 123/300\n",
      "34/34 - 3s - loss: 0.1327 - val_loss: 0.1930\n",
      "Epoch 124/300\n",
      "34/34 - 3s - loss: 0.1327 - val_loss: 0.1931\n",
      "Epoch 125/300\n",
      "34/34 - 2s - loss: 0.1326 - val_loss: 0.1931\n",
      "Epoch 126/300\n",
      "34/34 - 3s - loss: 0.1327 - val_loss: 0.1929\n",
      "Epoch 127/300\n",
      "34/34 - 2s - loss: 0.1325 - val_loss: 0.1930\n",
      "Epoch 128/300\n",
      "34/34 - 3s - loss: 0.1325 - val_loss: 0.1928\n",
      "Epoch 129/300\n",
      "34/34 - 3s - loss: 0.1327 - val_loss: 0.1929\n",
      "Epoch 130/300\n",
      "34/34 - 2s - loss: 0.1326 - val_loss: 0.1931\n",
      "Epoch 131/300\n",
      "34/34 - 2s - loss: 0.1327 - val_loss: 0.1930\n",
      "Epoch 132/300\n",
      "34/34 - 2s - loss: 0.1324 - val_loss: 0.1930\n",
      "Epoch 133/300\n",
      "34/34 - 3s - loss: 0.1323 - val_loss: 0.1930\n",
      "Epoch 134/300\n",
      "34/34 - 3s - loss: 0.1327 - val_loss: 0.1932\n",
      "Epoch 135/300\n",
      "34/34 - 2s - loss: 0.1326 - val_loss: 0.1929\n",
      "Epoch 136/300\n",
      "\n",
      "Epoch 00136: ReduceLROnPlateau reducing learning rate to 1.0000000656873453e-06.\n",
      "34/34 - 2s - loss: 0.1326 - val_loss: 0.1929\n",
      "Epoch 137/300\n",
      "34/34 - 3s - loss: 0.1324 - val_loss: 0.1929\n",
      "Epoch 138/300\n",
      "34/34 - 2s - loss: 0.1325 - val_loss: 0.1929\n",
      "Epoch 139/300\n",
      "34/34 - 2s - loss: 0.1324 - val_loss: 0.1930\n",
      "Epoch 140/300\n",
      "34/34 - 2s - loss: 0.1324 - val_loss: 0.1930\n",
      "Epoch 141/300\n",
      "34/34 - 3s - loss: 0.1322 - val_loss: 0.1930\n",
      "Epoch 142/300\n",
      "34/34 - 2s - loss: 0.1324 - val_loss: 0.1930\n",
      "Epoch 143/300\n",
      "34/34 - 3s - loss: 0.1322 - val_loss: 0.1930\n",
      "Epoch 144/300\n",
      "34/34 - 2s - loss: 0.1324 - val_loss: 0.1930\n",
      "Epoch 145/300\n",
      "34/34 - 3s - loss: 0.1321 - val_loss: 0.1930\n",
      "Epoch 146/300\n",
      "\n",
      "Epoch 00146: ReduceLROnPlateau reducing learning rate to 1.0000001111620805e-07.\n",
      "34/34 - 3s - loss: 0.1325 - val_loss: 0.1930\n",
      "Epoch 147/300\n",
      "34/34 - 2s - loss: 0.1323 - val_loss: 0.1930\n",
      "Epoch 148/300\n",
      "34/34 - 2s - loss: 0.1324 - val_loss: 0.1930\n",
      "Min validation loss history=0.1928359568119049\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(\n",
    "    trn_inputs, trn_y,\n",
    "    validation_data = (val_inputs, val_y),\n",
    "    batch_size=64,\n",
    "    epochs=300,\n",
    "    sample_weight=sample_weight,\n",
    "    callbacks=[\n",
    "        tf.keras.callbacks.ReduceLROnPlateau(verbose=1, monitor='val_loss'),\n",
    "        tf.keras.callbacks.ModelCheckpoint(f'model.h5',save_best_only=True, verbose=0, monitor='val_loss'),\n",
    "        tf.keras.callbacks.EarlyStopping(\n",
    "            patience=20, \n",
    "            monitor='val_loss',\n",
    "            verbose=0,\n",
    "            mode=\"auto\",\n",
    "            baseline=None,\n",
    "            restore_best_weights=True,\n",
    "        ),\n",
    "    ],\n",
    "    verbose=2\n",
    ")\n",
    "print(f\"Min validation loss history={min(history.history['val_loss'])}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:31:49.535168Z",
     "iopub.status.busy": "2020-10-22T19:31:49.533862Z",
     "iopub.status.idle": "2020-10-22T19:31:49.555366Z",
     "shell.execute_reply": "2020-10-22T19:31:49.554763Z"
    },
    "papermill": {
     "duration": 0.272252,
     "end_time": "2020-10-22T19:31:49.555504",
     "exception": false,
     "start_time": "2020-10-22T19:31:49.283252",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model.load_weights(f'model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:31:50.056083Z",
     "iopub.status.busy": "2020-10-22T19:31:50.054786Z",
     "iopub.status.idle": "2020-10-22T19:31:51.537220Z",
     "shell.execute_reply": "2020-10-22T19:31:51.536184Z"
    },
    "papermill": {
     "duration": 1.713587,
     "end_time": "2020-10-22T19:31:51.537353",
     "exception": false,
     "start_time": "2020-10-22T19:31:49.823766",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: shape=(), dtype=float32, numpy=0.19283596>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val_preds = model.predict(val_inputs)\n",
    "tf.reduce_mean(mcrmse(val_y, val_preds))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.345031,
     "end_time": "2020-10-22T19:31:52.125291",
     "exception": false,
     "start_time": "2020-10-22T19:31:51.780260",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Prepare submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:31:52.708783Z",
     "iopub.status.busy": "2020-10-22T19:31:52.707775Z",
     "iopub.status.idle": "2020-10-22T19:31:55.122359Z",
     "shell.execute_reply": "2020-10-22T19:31:55.120725Z"
    },
    "papermill": {
     "duration": 2.672047,
     "end_time": "2020-10-22T19:31:55.122487",
     "exception": false,
     "start_time": "2020-10-22T19:31:52.450440",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model_public = build_model(107, 107)\n",
    "model_private = build_model(130, 130)\n",
    "\n",
    "model_public.load_weights(f'model.h5')\n",
    "model_private.load_weights(f'model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:31:55.574172Z",
     "iopub.status.busy": "2020-10-22T19:31:55.572840Z",
     "iopub.status.idle": "2020-10-22T19:31:55.574935Z",
     "shell.execute_reply": "2020-10-22T19:31:55.575488Z"
    },
    "papermill": {
     "duration": 0.231893,
     "end_time": "2020-10-22T19:31:55.575653",
     "exception": false,
     "start_time": "2020-10-22T19:31:55.343760",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "public_inputs = [public_ohe, public_adj, public_bpps, public_bpps_stats,]\n",
    "private_inputs = [private_ohe, private_adj, private_bpps, private_bpps_stats,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:31:56.042067Z",
     "iopub.status.busy": "2020-10-22T19:31:56.040683Z",
     "iopub.status.idle": "2020-10-22T19:32:01.736093Z",
     "shell.execute_reply": "2020-10-22T19:32:01.735043Z"
    },
    "papermill": {
     "duration": 5.930932,
     "end_time": "2020-10-22T19:32:01.736239",
     "exception": false,
     "start_time": "2020-10-22T19:31:55.805307",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((629, 107, 5), (3005, 130, 5))"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_preds = [model_public.predict(public_inputs), model_private.predict(private_inputs)]\n",
    "test_dfs = [public_df, private_df]\n",
    "\n",
    "test_preds[0].shape, test_preds[1].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:32:02.198688Z",
     "iopub.status.busy": "2020-10-22T19:32:02.197712Z",
     "iopub.status.idle": "2020-10-22T19:32:07.904769Z",
     "shell.execute_reply": "2020-10-22T19:32:07.904182Z"
    },
    "papermill": {
     "duration": 5.939866,
     "end_time": "2020-10-22T19:32:07.904900",
     "exception": false,
     "start_time": "2020-10-22T19:32:01.965034",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "629it [00:01, 574.85it/s] \n",
      "3005it [00:02, 1106.20it/s]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((3634, 7), (457953, 6))"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds_ls = []\n",
    "for df, preds in zip(test_dfs, test_preds):\n",
    "    for i, uid in tqdm(enumerate(df.id)):\n",
    "        single_pred = preds[i]\n",
    "        single_df = pd.DataFrame(single_pred, columns=pred_cols)\n",
    "        single_df['id_seqpos'] = [f'{uid}_{x}' for x in range(single_df.shape[0])]\n",
    "        preds_ls.append(single_df)\n",
    "preds_df = pd.concat(preds_ls).groupby('id_seqpos').mean().reset_index()\n",
    "\n",
    "test_df.shape, preds_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:32:08.388324Z",
     "iopub.status.busy": "2020-10-22T19:32:08.387254Z",
     "iopub.status.idle": "2020-10-22T19:32:13.686409Z",
     "shell.execute_reply": "2020-10-22T19:32:13.687429Z"
    },
    "papermill": {
     "duration": 5.550466,
     "end_time": "2020-10-22T19:32:13.687663",
     "exception": false,
     "start_time": "2020-10-22T19:32:08.137197",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "wrote to submission.csv\n"
     ]
    }
   ],
   "source": [
    "submission = preds_df[['id_seqpos', 'reactivity', 'deg_Mg_pH10', 'deg_pH10', 'deg_Mg_50C', 'deg_50C']]\n",
    "submission.to_csv(f'submission.csv', index=False)\n",
    "print(f'wrote to submission.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2020-10-22T19:32:14.418561Z",
     "iopub.status.busy": "2020-10-22T19:32:14.417464Z",
     "iopub.status.idle": "2020-10-22T19:32:15.018935Z",
     "shell.execute_reply": "2020-10-22T19:32:15.018334Z"
    },
    "papermill": {
     "duration": 0.9735,
     "end_time": "2020-10-22T19:32:15.019052",
     "exception": false,
     "start_time": "2020-10-22T19:32:14.045552",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((457953, 6), (457953, 6), (3634, 7))"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "submission.shape, pd.read_csv(sample_sub_path).shape, test_df.shape"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  },
  "papermill": {
   "duration": 496.223639,
   "end_time": "2020-10-22T19:32:16.889954",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2020-10-22T19:24:00.666315",
   "version": "2.1.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}