Machine Learning on Snowflake.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "88e9b98b-2b4a-460a-9ed1-46bbf06a9540",
   "metadata": {},
   "outputs": [],
   "source": [
    "from snowflake.snowpark.session import Session\n",
    "import snowflake.snowpark.functions as F\n",
    "import snowflake.snowpark.types as T\n",
    "from snowflake.snowpark.window import Window\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "import sys\n",
    "import getpass\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "48e012ec-e9c4-41ff-8c69-7c57a93014bb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "connection successful!\n"
     ]
    }
   ],
   "source": [
    "from connector import sfcon\n",
    "\n",
    "session = sfcon.connect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3f6beaff-1a44-4e8e-a0a3-b6a90e2a78a3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Row(status='Statement executed successfully.')]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "session.sql('USE DATABASE HOL_DB').collect()\n",
    "session.sql('USE schema PUBLIC').collect()\n",
    "session.sql('USE warehouse HOL_WH').collect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "ac2e2b36-8cfa-4ee5-b951-e15608eecd9e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------------------------------------------------------------------------------------------------------------------------------------\n",
      "|\"created_on\"                      |\"name\"     |\"is_default\"  |\"is_current\"  |\"origin\"                 |\"owner\"         |\"comment\"  |\"options\"  |\"retention_time\"  |\n",
      "--------------------------------------------------------------------------------------------------------------------------------------------------------------------\n",
      "|2022-12-18 06:02:51.495000-08:00  |HOL_DB     |N             |Y             |                         |ACCOUNTADMIN    |           |           |1                 |\n",
      "|2022-12-11 04:04:59.800000-08:00  |ML_E2E     |N             |N             |                         |DATA_SCIENTIST  |           |           |1                 |\n",
      "|2022-11-15 19:29:04.242000-08:00  |SNOWFLAKE  |N             |N             |SNOWFLAKE.ACCOUNT_USAGE  |                |           |           |1                 |\n",
      "--------------------------------------------------------------------------------------------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "session.sql(\"show databases\").show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "970eebf6-f989-450c-a530-2691683bf80b",
   "metadata": {},
   "source": [
    "In this article, we will learn how to use snowpark python api for\n",
    " - Data Preprocessing\n",
    " - Feature engineering\n",
    " - Build machine learning model and,\n",
    " - Save trained models as UDFs (User defined functions) in snowflake\n",
    " \n",
    " Let's begins"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "7a862f1c-ca25-4044-82db-e394b3c388bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"data/Covid Data.csv.zip\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "9b4285a2-c3e5-49f1-a311-11197e6ca3a4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>USMER</th>\n",
       "      <th>MEDICAL_UNIT</th>\n",
       "      <th>SEX</th>\n",
       "      <th>PATIENT_TYPE</th>\n",
       "      <th>DATE_DIED</th>\n",
       "      <th>INTUBED</th>\n",
       "      <th>PNEUMONIA</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PREGNANT</th>\n",
       "      <th>DIABETES</th>\n",
       "      <th>...</th>\n",
       "      <th>ASTHMA</th>\n",
       "      <th>INMSUPR</th>\n",
       "      <th>HIPERTENSION</th>\n",
       "      <th>OTHER_DISEASE</th>\n",
       "      <th>CARDIOVASCULAR</th>\n",
       "      <th>OBESITY</th>\n",
       "      <th>RENAL_CHRONIC</th>\n",
       "      <th>TOBACCO</th>\n",
       "      <th>CLASIFFICATION_FINAL</th>\n",
       "      <th>ICU</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>03/05/2020</td>\n",
       "      <td>97</td>\n",
       "      <td>1</td>\n",
       "      <td>65</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>03/06/2020</td>\n",
       "      <td>97</td>\n",
       "      <td>1</td>\n",
       "      <td>72</td>\n",
       "      <td>97</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>09/06/2020</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>55</td>\n",
       "      <td>97</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>12/06/2020</td>\n",
       "      <td>97</td>\n",
       "      <td>2</td>\n",
       "      <td>53</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>21/06/2020</td>\n",
       "      <td>97</td>\n",
       "      <td>2</td>\n",
       "      <td>68</td>\n",
       "      <td>97</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   USMER  MEDICAL_UNIT  SEX  PATIENT_TYPE   DATE_DIED  INTUBED  PNEUMONIA  \\\n",
       "0      2             1    1             1  03/05/2020       97          1   \n",
       "1      2             1    2             1  03/06/2020       97          1   \n",
       "2      2             1    2             2  09/06/2020        1          2   \n",
       "3      2             1    1             1  12/06/2020       97          2   \n",
       "4      2             1    2             1  21/06/2020       97          2   \n",
       "\n",
       "   AGE  PREGNANT  DIABETES  ...  ASTHMA  INMSUPR  HIPERTENSION  OTHER_DISEASE  \\\n",
       "0   65         2         2  ...       2        2             1              2   \n",
       "1   72        97         2  ...       2        2             1              2   \n",
       "2   55        97         1  ...       2        2             2              2   \n",
       "3   53         2         2  ...       2        2             2              2   \n",
       "4   68        97         1  ...       2        2             1              2   \n",
       "\n",
       "   CARDIOVASCULAR  OBESITY  RENAL_CHRONIC  TOBACCO  CLASIFFICATION_FINAL  ICU  \n",
       "0               2        2              2        2                     3   97  \n",
       "1               2        1              1        2                     5   97  \n",
       "2               2        2              2        2                     3    2  \n",
       "3               2        2              2        2                     7   97  \n",
       "4               2        2              2        2                     3   97  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "34309706-ece0-4054-82c9-a25c98a1d7a5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1048575, 21)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1cff733f-d222-42f0-be6b-318a776dbc2d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1048575 entries, 0 to 1048574\n",
      "Data columns (total 21 columns):\n",
      " #   Column                Non-Null Count    Dtype \n",
      "---  ------                --------------    ----- \n",
      " 0   USMER                 1048575 non-null  int64 \n",
      " 1   MEDICAL_UNIT          1048575 non-null  int64 \n",
      " 2   SEX                   1048575 non-null  int64 \n",
      " 3   PATIENT_TYPE          1048575 non-null  int64 \n",
      " 4   DATE_DIED             1048575 non-null  object\n",
      " 5   INTUBED               1048575 non-null  int64 \n",
      " 6   PNEUMONIA             1048575 non-null  int64 \n",
      " 7   AGE                   1048575 non-null  int64 \n",
      " 8   PREGNANT              1048575 non-null  int64 \n",
      " 9   DIABETES              1048575 non-null  int64 \n",
      " 10  COPD                  1048575 non-null  int64 \n",
      " 11  ASTHMA                1048575 non-null  int64 \n",
      " 12  INMSUPR               1048575 non-null  int64 \n",
      " 13  HIPERTENSION          1048575 non-null  int64 \n",
      " 14  OTHER_DISEASE         1048575 non-null  int64 \n",
      " 15  CARDIOVASCULAR        1048575 non-null  int64 \n",
      " 16  OBESITY               1048575 non-null  int64 \n",
      " 17  RENAL_CHRONIC         1048575 non-null  int64 \n",
      " 18  TOBACCO               1048575 non-null  int64 \n",
      " 19  CLASIFFICATION_FINAL  1048575 non-null  int64 \n",
      " 20  ICU                   1048575 non-null  int64 \n",
      "dtypes: int64(20), object(1)\n",
      "memory usage: 168.0+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e0fc700-3cea-441f-b710-89dd5df0d4a9",
   "metadata": {},
   "source": [
    "```\n",
    "Dataset description\n",
    "\n",
    "sex: 1 for female and 2 for male.\n",
    "age: of the patient.\n",
    "classification: covid test findings. Values 1-3 mean that the patient was diagnosed with covid in different\n",
    "degrees. 4 or higher means that the patient is not a carrier of covid or that the test is inconclusive.\n",
    "patient type: type of care the patient received in the unit. 1 for returned home and 2 for hospitalization.\n",
    "pneumonia: whether the patient already have air sacs inflammation or not.\n",
    "pregnancy: whether the patient is pregnant or not.\n",
    "diabetes: whether the patient has diabetes or not.\n",
    "copd: Indicates whether the patient has Chronic obstructive pulmonary disease or not.\n",
    "asthma: whether the patient has asthma or not.\n",
    "inmsupr: whether the patient is immunosuppressed or not.\n",
    "hypertension: whether the patient has hypertension or not.\n",
    "cardiovascular: whether the patient has heart or blood vessels related disease.\n",
    "renal chronic: whether the patient has chronic renal disease or not.\n",
    "other disease: whether the patient has other disease or not.\n",
    "obesity: whether the patient is obese or not.\n",
    "tobacco: whether the patient is a tobacco user.\n",
    "usmr: Indicates whether the patient treated medical units of the first, second or third level.\n",
    "medical unit: type of institution of the National Health System that provided the care.\n",
    "intubed: whether the patient was connected to the ventilator.\n",
    "icu: Indicates whether the patient had been admitted to an Intensive Care Unit.\n",
    "date died: If the patient died indicate the date of death, and 9999-99-99 otherwise.\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "5dd2ff4f-1679-4d1a-ad3b-bb3ead159a3a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<snowflake.snowpark.table.Table at 0x28a2be58040>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#let's write dataset into snowflake\n",
    "session.write_pandas(df, table_name='COVID19_RECORDS', auto_create_table=True, overwrite=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b907a945-4956-4af5-bf2b-aa6590716621",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------\n",
      "|\"created_on\"                      |\"name\"                         |\"database_name\"  |\"schema_name\"  |\"kind\"  |\"comment\"  |\"cluster_by\"  |\"rows\"   |\"bytes\"   |\"owner\"       |\"retention_time\"  |\"automatic_clustering\"  |\"change_tracking\"  |\"is_external\"  |\n",
      "--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------\n",
      "|2022-12-18 06:04:48.640000-08:00  |APPLICATION_RECORD             |HOL_DB           |PUBLIC         |TABLE   |           |              |438557   |3974656   |ACCOUNTADMIN  |1                 |OFF                     |OFF                |N              |\n",
      "|2022-12-25 21:50:29.809000-08:00  |COVID19_RECORDS                |HOL_DB           |PUBLIC         |TABLE   |           |              |1048575  |10075136  |ACCOUNTADMIN  |1                 |OFF                     |OFF                |N              |\n",
      "|2022-12-25 21:08:01.410000-08:00  |COVID19_RECORDS_PROCESSED      |HOL_DB           |PUBLIC         |TABLE   |           |              |78188    |492032    |ACCOUNTADMIN  |1                 |OFF                     |OFF                |N              |\n",
      "|2022-12-18 06:05:13.851000-08:00  |CREDIT_RECORD                  |HOL_DB           |PUBLIC         |TABLE   |           |              |1048575  |1348608   |ACCOUNTADMIN  |1                 |OFF                     |OFF                |N              |\n",
      "|2022-12-18 08:26:34.800000-08:00  |CREDIT_RISK_PREPARED           |HOL_DB           |PUBLIC         |TABLE   |           |              |30294    |228864    |ACCOUNTADMIN  |1                 |OFF                     |OFF                |N              |\n",
      "|2022-12-18 08:35:11.779000-08:00  |CREDIT_RISK_PREPARED_BALANCED  |HOL_DB           |PUBLIC         |TABLE   |           |              |59616    |549376    |ACCOUNTADMIN  |1                 |OFF                     |OFF                |N              |\n",
      "|2022-12-18 06:35:43.103000-08:00  |MY_FIRST_ANALYSIS              |HOL_DB           |PUBLIC         |TABLE   |           |              |438510   |5106176   |ACCOUNTADMIN  |1                 |OFF                     |OFF                |N              |\n",
      "--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#check tables available in current sechama\n",
    "session.sql('show tables').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e8bfe627-dbef-457c-8f37-ea555ca08376",
   "metadata": {},
   "source": [
    "# Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "037ca14b-c660-4d9a-a426-3f13c037c433",
   "metadata": {},
   "outputs": [],
   "source": [
    "snowpark_df = session.table(\"COVID19_RECORDS\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e23f9d40-0d22-4ea3-95fc-bec3f65cf9f3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "snowflake.snowpark.table.Table"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(snowpark_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f9ab5574-3e74-419b-8a94-f55ff562258e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size of the table object: 4.8e-05 MB\n"
     ]
    }
   ],
   "source": [
    "print(f\"Size of the table object: {(sys.getsizeof(snowpark_df)/1e6)} MB\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "4cfcb0f0-b9f6-438a-b3bd-62db5b4ac185",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[1;31mInit signature:\u001b[0m\n",
       "\u001b[0msnowflake\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msnowpark\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtable\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTable\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m\n",
       "\u001b[0m    \u001b[0mtable_name\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mstr\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\n",
       "\u001b[0m    \u001b[0msession\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mUnion\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mForwardRef\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'snowflake.snowpark.session.Session'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mNoneType\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\n",
       "\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m->\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
       "\u001b[1;31mDocstring:\u001b[0m     \n",
       "Represents a lazily-evaluated Table. It extends :class:`DataFrame` so all\n",
       ":class:`DataFrame` operations can be applied to it.\n",
       "\n",
       "You can create a :class:`Table` object by calling :meth:`Session.table`\n",
       "with the name of the table in Snowflake. See examples in :meth:`Session.table`.\n",
       "\u001b[1;31mFile:\u001b[0m           c:\\users\\ransaka_09914\\anaconda3\\envs\\getting_started_snowpark_python\\lib\\site-packages\\snowflake\\snowpark\\table.py\n",
       "\u001b[1;31mType:\u001b[0m           type\n",
       "\u001b[1;31mSubclasses:\u001b[0m     \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import snowflake\n",
    "?snowflake.snowpark.table.Table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "67bbf344-45f5-43b7-8c70-5e2177eb361d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1048575"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "snowpark_df.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f8606bc6-d244-4a40-b165-77d3df188898",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(snowpark_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "b0c76e7a-9c26-4de9-bb40-c6cc72e2d6d2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[StructField('USMER', LongType(), nullable=True),\n",
       " StructField('MEDICAL_UNIT', LongType(), nullable=True),\n",
       " StructField('SEX', LongType(), nullable=True),\n",
       " StructField('PATIENT_TYPE', LongType(), nullable=True),\n",
       " StructField('DATE_DIED', StringType(), nullable=True),\n",
       " StructField('INTUBED', LongType(), nullable=True),\n",
       " StructField('PNEUMONIA', LongType(), nullable=True),\n",
       " StructField('AGE', LongType(), nullable=True),\n",
       " StructField('PREGNANT', LongType(), nullable=True),\n",
       " StructField('DIABETES', LongType(), nullable=True),\n",
       " StructField('COPD', LongType(), nullable=True),\n",
       " StructField('ASTHMA', LongType(), nullable=True),\n",
       " StructField('INMSUPR', LongType(), nullable=True),\n",
       " StructField('HIPERTENSION', LongType(), nullable=True),\n",
       " StructField('OTHER_DISEASE', LongType(), nullable=True),\n",
       " StructField('CARDIOVASCULAR', LongType(), nullable=True),\n",
       " StructField('OBESITY', LongType(), nullable=True),\n",
       " StructField('RENAL_CHRONIC', LongType(), nullable=True),\n",
       " StructField('TOBACCO', LongType(), nullable=True),\n",
       " StructField('CLASIFFICATION_FINAL', LongType(), nullable=True),\n",
       " StructField('ICU', LongType(), nullable=True)]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "snowpark_df.schema.fields"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "69a7f915-1b46-4f51-9b23-f9d2f7912931",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Column[\"USMER\"] has 2 unique values\n",
      "Column[\"MEDICAL_UNIT\"] has 13 unique values\n",
      "Column[\"SEX\"] has 2 unique values\n",
      "Column[\"PATIENT_TYPE\"] has 2 unique values\n",
      "Column[\"DATE_DIED\"] has 401 unique values\n",
      "Column[\"INTUBED\"] has 4 unique values\n",
      "Column[\"PNEUMONIA\"] has 3 unique values\n",
      "Column[\"AGE\"] has 121 unique values\n",
      "Column[\"PREGNANT\"] has 4 unique values\n",
      "Column[\"DIABETES\"] has 3 unique values\n",
      "Column[\"COPD\"] has 3 unique values\n",
      "Column[\"ASTHMA\"] has 3 unique values\n",
      "Column[\"INMSUPR\"] has 3 unique values\n",
      "Column[\"HIPERTENSION\"] has 3 unique values\n",
      "Column[\"OTHER_DISEASE\"] has 3 unique values\n",
      "Column[\"CARDIOVASCULAR\"] has 3 unique values\n",
      "Column[\"OBESITY\"] has 3 unique values\n",
      "Column[\"RENAL_CHRONIC\"] has 3 unique values\n",
      "Column[\"TOBACCO\"] has 3 unique values\n",
      "Column[\"CLASIFFICATION_FINAL\"] has 7 unique values\n",
      "Column[\"ICU\"] has 4 unique values\n"
     ]
    }
   ],
   "source": [
    "for column in snowpark_df:\n",
    "    print(column,'has',snowpark_df[[column]].distinct().count(),'unique values')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "cb04b5ce-c10b-47d8-a12f-dbee5a4fa079",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------\n",
      "|\"SEX\"  |\n",
      "---------\n",
      "|1      |\n",
      "|2      |\n",
      "|2      |\n",
      "|1      |\n",
      "|2      |\n",
      "|1      |\n",
      "|1      |\n",
      "|1      |\n",
      "|1      |\n",
      "|1      |\n",
      "---------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "snowpark_df.select(\"SEX\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "2932e7d9-d0af-4f73-aee6-a005284091ac",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------\n",
      "|\"SUMMARY\"  |\"CLASIFFICATION_FINAL\"  |\n",
      "--------------------------------------\n",
      "|count      |1048575.0               |\n",
      "|mean       |5.305653                |\n",
      "|stddev     |1.8811653303205436      |\n",
      "|max        |7.0                     |\n",
      "|min        |1.0                     |\n",
      "--------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "snowpark_df[['CLASIFFICATION_FINAL']].describe().show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "bac5108d-1530-401a-a137-08f38c34d761",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "------------------------------------\n",
      "|\"CLASIFFICATION_FINAL\"  |\"COUNT\"  |\n",
      "------------------------------------\n",
      "|3                       |381527   |\n",
      "|5                       |26091    |\n",
      "|7                       |499250   |\n",
      "|6                       |128133   |\n",
      "|1                       |8601     |\n",
      "|2                       |1851     |\n",
      "|4                       |3122     |\n",
      "------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "snowpark_df.group_by(\"CLASIFFICATION_FINAL\").count().show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc075605-f069-49a7-a730-d0b1de78d47a",
   "metadata": {},
   "source": [
    "```sql \n",
    "SELECT\n",
    "    \"USMER\",\n",
    "    \"MEDICAL_UNIT\",\n",
    "    \"SEX\",\n",
    "    \"PATIENT_TYPE\",\n",
    "    \"DATE_DIED\",\n",
    "    \"INTUBED\",\n",
    "    \"PNEUMONIA\",\n",
    "    \"AGE\",\n",
    "    \"PREGNANT\",\n",
    "    \"DIABETES\",\n",
    "    \"COPD\",\n",
    "    \"ASTHMA\",\n",
    "    \"INMSUPR\",\n",
    "    \"HIPERTENSION\",\n",
    "    \"OTHER_DISEASE\",\n",
    "    \"CARDIOVASCULAR\",\n",
    "    \"OBESITY\",\n",
    "    \"RENAL_CHRONIC\",\n",
    "    \"TOBACCO\",\n",
    "    \"CLASIFFICATION_FINAL\",\n",
    "    \"ICU\",\n",
    "    CASE\n",
    "        WHEN (\"CLASIFFICATION_FINAL\" < 4 :: INT) THEN 1 :: INT\n",
    "        ELSE 0 :: INT\n",
    "    END AS \"TARGET\"\n",
    "FROM\n",
    "    COVID19_RECORDS\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "3492f1b7-acbb-4082-9c0e-1041ec7f6368",
   "metadata": {},
   "outputs": [],
   "source": [
    "snowpark_df = snowpark_df.with_column('TARGET', F.when(F.col('CLASIFFICATION_FINAL')\n",
    "                        < 4, 1).otherwise(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "cfdb3b63-bfda-4e59-bd3f-1a2022c51ec4",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "snowpark_df\\\n",
    ".group_by(\"TARGET\").count().to_pandas().set_index(\"TARGET\")\\\n",
    ".plot.bar()\n",
    "\n",
    "plt.title(\"Target distribution\",fontweight='semibold')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "f5baebdf-59e4-4cf6-8827-a8d94b32670a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Input key or value type doesn't match the target column data type, this replacement was skipped. Column Name: \"DATE_DIED\", Type: StringType(), Input Key: 97, Type: <class 'int'>, Input Value: None, Type: <class 'NoneType'>\n",
      "Input key or value type doesn't match the target column data type, this replacement was skipped. Column Name: \"DATE_DIED\", Type: StringType(), Input Key: 99, Type: <class 'int'>, Input Value: None, Type: <class 'NoneType'>\n"
     ]
    }
   ],
   "source": [
    "snowpark_df = snowpark_df.replace([97,99],value=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "c7182621-227e-409e-aa00-e2d28e31d494",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "snowpark_df\\\n",
    ".select('AGE').to_pandas()\\\n",
    ".plot.hist(bins=100,alpha=0.8)\n",
    "\n",
    "plt.title(\"Age distribution\",fontweight='semibold')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "17e7af17-aa97-42fd-97b3-c36a57e90e54",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------------------\n",
      "|\"OO\"       |\"AGE\"  |\n",
      "---------------------\n",
      "|OLD ADULT  |65     |\n",
      "|OLD ADULT  |72     |\n",
      "|OLD ADULT  |55     |\n",
      "|OLD ADULT  |53     |\n",
      "|OLD ADULT  |68     |\n",
      "|ADULT      |40     |\n",
      "|OLD ADULT  |64     |\n",
      "|OLD ADULT  |64     |\n",
      "|ADULT      |37     |\n",
      "|ADULT      |25     |\n",
      "---------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "snowpark_df.select(F.when(F.col('AGE') < 21, 'YOUNG'\n",
    "                   ).otherwise(F.when(F.col('AGE') < 49, 'ADULT'\n",
    "                   ).otherwise('OLD ADULT')).as_('OO'), 'AGE').show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "6fe97f7f-48d0-42b3-afaa-bbb5646c506c",
   "metadata": {},
   "outputs": [],
   "source": [
    "snowpark_df = snowpark_df.with_column(\n",
    "    \"AGE_BKT\",\n",
    "    F.when(F.col(\"AGE\") < 21, \"YOUNG\").otherwise(\n",
    "        F.when(F.col(\"AGE\") < 49, \"ADULT\").otherwise(\"OLD ADULT\")\n",
    "    ),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "d759eb47-1aff-4db3-aa8d-a1172174433a",
   "metadata": {},
   "outputs": [],
   "source": [
    "age_bkt_df = snowpark_df.select(\n",
    "    F.col(\"AGE_BKT\"),\n",
    "    F.when((F.col(\"AGE_BKT\")=='YOUNG') & (F.col(\"TARGET\")==1),1).otherwise(0).as_(\"YOUNG_\"),\n",
    "    F.when((F.col(\"AGE_BKT\")=='ADULT') & (F.col(\"TARGET\")==1),1).otherwise(0).as_(\"ADULT_\"),\n",
    "    F.when((F.col(\"AGE_BKT\")=='OLD ADULT') & (F.col(\"TARGET\")==1),1).otherwise(0).as_(\"OLD_ADULT_\")\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b5a1cf44-f56b-4c8e-8090-b2f0bae6e390",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-----------------------\n",
      "|\"AGE_BKT\"  |\"COUNT\"  |\n",
      "-----------------------\n",
      "|OLD ADULT  |342413   |\n",
      "|ADULT      |628554   |\n",
      "|YOUNG      |77608    |\n",
      "-----------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "age_bkt_df.group_by(F.col(\"AGE_BKT\")).count().show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "c3fb745d-4bc5-4c4a-b8cb-2a9c976a4500",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "age_bkt_df.group_by(F.col(\"AGE_BKT\")).count().to_pandas().set_index(\"AGE_BKT\")\\\n",
    ".plot.barh(alpha=0.8)\n",
    "\n",
    "plt.title(\"Cases count by gender bucket\",fontweight='semibold')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "f7cfeabd-15a6-412b-aa52-7743f4ac6cdd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------------------------------------\n",
      "|\"YOUNG % OF CASES\"  |\"ADULT % OF CASES\"  |\"OLD_ADULT % OF CASES\"  |\n",
      "--------------------------------------------------------------------\n",
      "|1.534463            |20.877858           |14.969745               |\n",
      "--------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "age_bkt_df.select(\n",
    "    ((F.sum(\"YOUNG_\") * 100 ) / F.count(\"YOUNG_\")).as_(\"YOUNG % OF CASES\"),\n",
    "    ((F.sum(\"ADULT_\") * 100) / F.count(\"ADULT_\")).as_(\"ADULT % OF CASES\"),\n",
    "    ((F.sum(\"OLD_ADULT_\") * 100) / F.count(\"OLD_ADULT_\")).as_(\"OLD_ADULT % OF CASES\")\n",
    ").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "f216f7e8-12f1-4374-bc84-79379b6472b4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArkAAAGdCAYAAAAWi2sJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA4BElEQVR4nO3dd3hVVf7+/Ts9QAqdJLRQUggoIiBNkRpAijqUBASBARUBlTYoMA6RjtLkpzRJKEpzBFSwUISgSFUCUkIRQhlNAJEkCAgh2c8fPDlfjycNMMYs36/r2tfI3muv/Vlncebc2Vln42RZliUAAADAIM4FXQAAAADwRyPkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADCOa0EXABSUjIwM/fjjj/L29paTk1NBlwMAAPLAsixduXJFAQEBcnbO/n4tIRd/Wz/++KMqVqxY0GUAAIC7cO7cOVWoUCHb44Rc/G15e3tLuv0m8fHxKeBqAABAXqSmpqpixYq2z/HsEHLxt5W5RMHHx4eQCwBAIZPbUkO+eAYAAADjEHIBAABgHEIuAAAAjMOaXAAAgELEsizdunVL6enpBV1KvnBxcZGrq+s9P96TkAsAAFBI3Lx5U4mJibp27VpBl5KvihYtKn9/f7m7u991H4RcAACAQiAjI0MJCQlycXFRQECA3N3djfvHjCzL0s2bN3Xx4kUlJCQoKCgox3/wISeEXAAAgELg5s2bysjIUMWKFVW0aNGCLiffFClSRG5ubjpz5oxu3rwpT0/Pu+qHL54BAAAUInd7Z7Mw+SPGaP6rBAAAgL8dQi4AAACMw5pcAACAQizwlU/+1OudntL+T73e3eJOLgAAAPLdnDlzVKVKFXl6eqpu3br66quv8vV6hFwAAADkq1WrVmnIkCEaM2aM4uLi9Mgjj6hdu3Y6e/Zsvl2TkAsAAIB8NWPGDPXr10/9+/dXjRo1NGvWLFWsWFFz587Nt2sScgEAAJBvbt68qW+//Vbh4eF2+8PDw7Vjx458uy4hFwAAAPnmp59+Unp6usqVK2e3v1y5ckpKSsq36xJyAQAAkO9+/08QW5aVr/8sMSEXAAAA+aZ06dJycXFxuGt74cIFh7u7fyRCLgAAAPKNu7u76tatq02bNtnt37Rpkxo3bpxv1+UfgwAAAEC+GjZsmHr16qV69eqpUaNGWrBggc6ePasBAwbk2zUJuQAAAIVYYfgXyCIiInTp0iWNGzdOiYmJqlWrlj799FNVrlw5365JyAUAAEC+GzhwoAYOHPinXY81uQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAUIhYllXQJeS7P2KMPF0Bf3u1xm6Qs0fRgi4DAAqlwvD4KlO4ublJkq5du6YiRYoUcDX569q1a5L+b8x3g5ALAABQCLi4uKh48eK6cOGCJKlo0aJycnIq4Kr+WJZl6dq1a7pw4YKKFy8uFxeXu+6LkAsAAFBI+Pn5SZIt6JqqePHitrHeLUIuAABAIeHk5CR/f3+VLVtWaWlpBV1OvnBzc7unO7iZCLkAAACFjIuLyx8SBE3G0xUAAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDh3HHLPnTunfv36KSAgQO7u7qpcubJeeuklXbp0ya5ds2bNNGTIkGz7cXJysm3FihVTUFCQ+vTpo2+//faOByFJISEhcnd31w8//OBwrFmzZrZreXh4qHz58urYsaPWrFlj1+706dNycnLS/v37Hfp44okn1KdPn2zHFxsbazemrLbFixdnWfuwYcNUsmRJVapUSStXrrQ79v7776tjx455eg2uX7+usWPHKiQkRB4eHipdurS6dOmiw4cP27WLiorKsr7Nmzfn2P/q1avVrFkz+fr6ysvLS/fff7/GjRunn3/+2aGOEiVKqGTJkrp+/bpDP3FxcerQoYPKli0rT09PBQYGKiIiQj/99JOk/5uHrLZdu3ZJktLT0zV58mSFhoaqSJEiKlmypBo2bKhFixbl6bUCAABmu6OQe+rUKdWrV0/Hjx/XihUr9P3332vevHn64osv1KhRI4ewk5tFixYpMTFRhw8f1ttvv61ffvlFDRo00NKlS++on+3bt+vXX39V165dsw2SzzzzjBITE/X9999r9erVCgsLU2RkpJ599tk7ulZ2GjdurMTERNvWrVs3tW3b1m5fRESEw3nr1q3T8uXLtXHjRk2dOlV9+/a1/cCQnJysMWPG6O233871+jdu3FCrVq0UExOj8ePH6/jx4/r000+Vnp6uBg0a2MJhppo1a9rVlpiYqKZNm2bb/5gxYxQREaH69evrs88+06FDhzR9+nQdOHBA7777rl3b1atXq1atWgoLC3P4QeLChQtq1aqVSpcurQ0bNig+Pl4xMTHy9/fXtWvX7Npu3rzZoca6detKuh3UZ82apfHjx+vIkSPaunWrnnnmGV2+fDnX1woAAJjP9U4aDxo0SO7u7tq4caOKFCkiSapUqZLq1KmjatWqacyYMZo7d26e+ytevLj8/PwkSYGBgQoPD1fv3r01ePBgdezYUSVKlMhTP9HR0erRo4ceffRRDRo0SKNHj5aTk5Ndm6JFi9quVbFiRTVs2FChoaH65z//qW7duqlVq1Z5rjsr7u7utv4lqUiRIrpx44bdvqzEx8erWbNmqlevnurVq6chQ4bo1KlTKlWqlEaOHKmBAweqUqVKuV5/1qxZ2rlzp+Li4lS7dm1JUuXKlbV69Wo1aNBA/fr106FDh2yvi6ura661ZdqzZ48mTZqkWbNm6aWXXrLtDwwMVOvWrZWcnGzXPjo6Wj179pRlWYqOjtZTTz1lO7Zjxw6lpqZq4cKFcnW9/devSpUqatGihcN1S5UqlW2N69at08CBA9W1a1fbvsxxAwAA5PlO7s8//6wNGzZo4MCBtoCbyc/PT0899ZRWrVoly7LuqaChQ4fqypUr2rRpU57aX7lyRf/973/Vs2dPtW7dWlevXlVsbGyezu3du7dKlCjhcLfxz1S7dm198803unz5sr799ltdv35d1atX1/bt27Vv3z69+OKLeepn+fLlat26tUPQc3Z21tChQ3XkyBEdOHDgrmpctmyZvLy8NHDgwCyPFy9e3PbfJ0+e1M6dO9WtWzd169ZNO3bs0KlTp2zH/fz8dOvWLa1du/ae/q74+flpy5YtunjxYp7PuXHjhlJTU+02AABgpjyH3BMnTsiyLNWoUSPL4zVq1NDly5fvKHRkJTQ0VNLtdZl5sXLlSgUFBalmzZpycXFRZGSkoqOj83Sus7OzgoOD83yt/NCmTRv17NlT9evXV58+fbRkyRIVK1ZMzz//vObPn6+5c+cqJCRETZo0cVhb+1vHjx/PcW4y22Q6ePCgvLy8bNtDDz2Ubd8nTpxQ1apV5ebmlut4YmJi1K5dO9ua3LZt2yomJsZ2vGHDhho9erR69Oih0qVLq127dnrjjTd0/vx5h74aN25sV6OXl5fS09MlSTNmzNDFixfl5+en+++/XwMGDNBnn32WY22TJ0+Wr6+vbatYsWKu4wEAAIXTH/Z0hcy7cr9fJpDf/WT+ajxTz549tWbNGodfoed0vXut+V5FRUXp+++/18GDB/Xkk09q0qRJatWqldzc3DRhwgRt375d/fv319NPP31X/Wf1moaEhGj//v22bfXq1Tmen5fXKD09XUuWLHGYjyVLltjCqSRNnDhRSUlJmjdvnsLCwjRv3jyFhobq4MGDdv2tWrXKrsb9+/fLxcVFkhQWFqZDhw5p165d6tu3r86fP6+OHTuqf//+2dY3atQopaSk2LZz587lOiYAAFA45TnkVq9eXU5OTjpy5EiWx48ePaoSJUqodOnS91RQfHy8pNvrNHNz5MgR7d69WyNHjpSrq6tcXV3VsGFDXb9+XStWrMj1/PT0dJ04ccJ2LV9fX0lSSkqKQ9vk5GTb8fx09OhRLVu2TOPHj1dsbKyaNm2qMmXKqFu3btq3b1+2v2IPDg7OcW4kKSgoyLbP3d1d1atXt2053dUMDg7WyZMnlZaWlmPtGzZs0A8//KCIiAjbfERGRup///ufNm7caNe2VKlS6tq1q6ZPn674+HgFBARo2rRpdm0qVqxoV2P16tXtjjs7O6t+/foaOnSo1q5dq8WLFys6OloJCQlZ1ufh4SEfHx+7DQAAmCnPIbdUqVJq3bq15syZ4/BYqKSkJC1btkwRERH3fFd01qxZ8vHxydMXwaKjo9W0aVMdOHDA7m7fyJEj87RkYcmSJbp8+bI6d+4sSSpRooTKlCmjvXv32rW7fv26Dh8+rJCQkLsbVB5ZlqVnn31W06dPt/1qPjNYZv5vRkZGludGRkZq8+bNDutuMzIyNHPmTIWFhd31F7N69OihX375RXPmzMnyeOZd8+joaEVGRjrcfX3qqadynA93d3dVq1ZNV69evav6MoWFhUnSPfcDAAAKvzt6usJbb72lxo0bq02bNpowYYKqVKmiw4cP61//+pfKly+viRMn2rW/ePGiwzNn/fz8bN+YT05OVlJSkm7cuKHjx49r/vz5+vDDD7V06VK7LzNlJS0tTe+++67GjRunWrVq2R3r37+/Xn/9dR04cMAW7K5du6akpCTdunVLP/zwg9asWaOZM2fq+eefV/PmzW3njhgxQpMmTVK5cuXUuHFjXb58WVOnTpWrq6vdr+HzMr479c4776hs2bLq1KmTJKlJkyaKiorSrl279NlnnyksLCzb12Xo0KH66KOP1LFjR02fPl0NGjTQ+fPnNWnSJMXHx2vz5s13/QNIgwYNNHLkSA0fPlw//PCDnnzySQUEBNgeIffwww+rR48eWrdunT7++GOH+ejdu7fat2+vixcvavfu3Vq5cqUiIyMVHBwsy7K0bt06ffrppw7PuL106ZKSkpLs9hUvXlyenp7q0qWLmjRposaNG8vPz08JCQkaNWqUgoODbeu6AQDA39cdhdygoCB98803ioqKUkREhC5duiQ/Pz898cQTGjt2rEqWLGnXfvny5Vq+fLndvrFjxyoqKkqS1LdvX0mSp6enypcvr4cfflh79uzRgw8+mGstH3/8sS5duqQnn3wyyzrvu+8+RUdHa/bs2ZJuB8h33nlH7u7uKlWqlOrWratVq1Y5nD9ixAh5eXlp2rRpOnnypIoXL66GDRvqq6++cvj1dm7juxOZgXTHjh22fQ899JCGDx+u9u3bq2zZslqyZEm253t6emrLli2aPHmyRo8erTNnzsjb21vNmzfXrl27HILnnZo6darq1q2rt99+W/PmzVNGRoaqVaumLl26qHfv3oqOjlaxYsXUsmVLh3ObN28ub29vvfvuu3riiSdUtGhRDR8+XOfOnZOHh4eCgoK0cOFC9erVy+68rO7mr1ixQpGRkWrTpo1WrFihyZMnKyUlRX5+fmrRooWioqJsjyYDAAB/X07WvT7zCyikUlNTbz9lYcj7cvYoWtDlAEChdHpK+4IuAX8zmZ/fKSkpOX6/5g97ugIAAADwV/GXDrnt2rVzeE5q5jZp0qSCLg8AAAB/UX/pxYsLFy50eJJDpt+v/wUAAAAy/aVDbvny5Qu6BAAAABRCf+nlCgAAAMDdIOQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjuBZ0AUBBO/RaG/n4+BR0GQAA4A/EnVwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHFcC7oAoKDVGrtBzh5FC7oMAACMcXpK+4IugTu5AAAAMA8hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMI4RIXfHjh1ycXFR27ZtHY6dPn1aTk5Ots3b21s1a9bUoEGDdOLECbu2UVFReuCBBxz6SE5OlpOTk2JjY+363L9/v61Nnz597K6T1ZaV06dPq2nTpvLy8tKjjz6qM2fO2B1v3769Vq9enafX4fDhw+rWrZvKlCkjDw8PBQUF6dVXX9W1a9fs2gUGBjrUVqFChRz7Tk1N1ZgxYxQaGipPT0/5+fmpVatWWrNmjSzLsmu7fPlyubi4aMCAAVn2NX/+fNWuXVvFihVT8eLFVadOHU2dOtV2PCoqKsvXLzQ01Nbm1KlT6t69uwICAuTp6akKFSro8ccf1/Hjx/P0WgEAALMZEXJjYmL0wgsvaPv27Tp79myWbTZv3qzExEQdOHBAkyZNUnx8vGrXrq0vvvjiD6nhzTffVGJiom2TpEWLFjns+73hw4erfPnyiouLk5+fn0aMGGE7tnLlSrm4uKhz5865Xn/Xrl1q0KCBbt68qU8++UTHjx/XpEmTtGTJErVu3Vo3b960az9u3Di72uLi4rLtOzk5WY0bN9bSpUs1atQo7du3T19++aUiIiI0cuRIpaSk2LWPiYnRyJEjtXLlSoeAHR0drWHDhunFF1/UgQMH9PXXX2vkyJH65Zdf7NrVrFnTrr7ExERt375dknTz5k21bt1aqampWrNmjY4dO6ZVq1apVq1aDrUAAIC/J9eCLuBeXb16Ve+//7727t2rpKQkLV68WP/5z38c2pUqVUp+fn6SpKpVq6pjx45q2bKl+vXrp5MnT8rFxeWe6vD19ZWvr6/dvuLFi9uumZ34+HjNmDFDQUFB6tOnjy3kJicn69///re2bNmS67Uty1K/fv1Uo0YNrVmzRs7Ot392qVy5soKDg1WnTh3NnDlTL7/8su0cb2/vXGvLNHr0aJ0+fVrHjx9XQECAbX9wcLC6d+8uT09P277Tp09rx44dWr16tbZu3aoPPvhATz/9tO34unXr1K1bN/Xr18+2r2bNmg7XdHV1zba+I0eO6NSpU9qyZYsqV65sG2uTJk3yNB4AAGC+Qn8nd9WqVQoJCVFISIh69uypRYsWOfz6PCvOzs566aWXdObMGX377bd/QqVZq127tjZv3qyMjAxt3LhR999/vyRpxIgRGjx4sCpVqpRrH/v379eRI0c0bNgwW8D9bf+tWrXSihUr7qq+jIwMrVy5Uk899ZRdwM3k5eUlV9f/+1kpJiZG7du3l6+vr3r27Kno6Gi79n5+ftq1a5fDsow7UaZMGTk7O+uDDz5Qenp6ns+7ceOGUlNT7TYAAGCmQh9yo6Oj1bNnT0lS27Zt9csvv+R5CULmGs/Tp0/nV3m5mjZtmo4eParAwECdOHFC06ZN05dffqkDBw7o6aefVrdu3VS1alUNGDDAYclBpsx1qDVq1MjyeI0aNRzWqr788svy8vKybbNnz87y3J9++kmXL1+2Ww+bnYyMDC1evNg2H5GRkdq5c6e+//57W5uxY8eqePHiCgwMVEhIiPr06aP3339fGRkZdn0dPHjQrj4vLy/1799fklS+fHnNnj1b//nPf1SiRAm1aNFC48eP16lTp3Ksb/LkybY77r6+vqpYsWKuYwIAAIVToQ65x44d0549exQZGSnp9q+4IyIiFBMTk6fzM+/4ZvelsD9D+fLltX79ep09e1br169X6dKlNXDgQM2fP18TJkyQt7e3jh07phMnTmj+/Pl3dQ3LshzG+K9//Uv79++3bb9dUvD7c6W8vUYbN27U1atX1a5dO0lS6dKlFR4ebjcf/v7+2rlzpw4ePKgXX3xRaWlp6t27t9q2bWsXdENCQuzq279/vyZOnGg7PmjQICUlJem9995To0aN9N///lc1a9bUpk2bsq1v1KhRSklJsW3nzp3LdUwAAKBwKtRrcqOjo3Xr1i2VL1/ets+yLLm5ueny5csqUaJEjufHx8dLkqpUqSJJ8vHxyfKLS8nJyZLksOY2P0ycOFHh4eF68MEH1b9/f02YMEFubm76xz/+oS1btuiFF15wOCc4OFjS7bWqWT0d4ujRowoKCrLbV7p0aVWvXj3XesqUKaMSJUrYXqucxMTE6Oeff1bRokVt+zIyMhQXF6fx48fbrXuuVauWatWqpUGDBmn79u165JFHtG3bNjVv3lyS5O7unmt93t7e6tSpkzp16qQJEyaoTZs2mjBhglq3bp1lew8PD3l4eOQ6DgAAUPgV2ju5t27d0tKlSzV9+nS7u30HDhxQ5cqVtWzZshzPz8jI0OzZs1WlShXVqVNH0u3lC//73/+UlJRk13bv3r1ydnbOUyi8F/Hx8VqxYoXGjRsnSUpPT1daWpokKS0tLdv1pw888IBCQ0M1c+ZMh1/7HzhwQJs3b1b37t3vqiZnZ2dFRERo2bJl+vHHHx2OX716Vbdu3dKlS5f00UcfaeXKlQ53YH/55Rd99tln2V4jLCzM1tfdynzE2L30AQAAzFFo7+SuX79ely9fVr9+/RzusHbp0kXR0dEaPHiwbd+lS5eUlJSka9eu6dChQ5o1a5b27NmjTz75xHaHMTw8XDVq1FBkZKQmTpyogIAAfffddxoxYoQGDBggb29vu+scO3bMoa6wsDC5u7vf8Xgsy9Kzzz6rmTNnysvLS5LUpEkTvfPOOwoODtbSpUuzDapOTk5auHChwsPD1blzZ40aNUp+fn7avXu3hg8frkaNGmnIkCF3XFOmSZMmKTY2Vg0aNNDEiRNVr149ubm56auvvtLkyZO1d+9evfvuuypVqpS6du3q8OW3Dh06KDo6Wh06dNDzzz+vgIAAtWjRQhUqVFBiYqImTJigMmXKqFGjRrZzbt265fDDhpOTk8qVK6f9+/dr7Nix6tWrl+313rZtm2JiYuyeIAEAAP6+Cm3IjY6OVqtWrbJcQtC5c2dNmjRJ+/btU8mSJSVJrVq1kiQVLVpUlStXVvPmzbVgwQK7u7Ourq7auHGjRo8eraeeekoXLlxQ5cqV1b9/f40cOdLhOplrgX8rISFBgYGBdzyeBQsWqFy5curQoYNtX1RUlHr06KEGDRqobdu2GjRoULbnN2nSRLt27dJrr72mxx57TKmpqapUqZJ69+6tUaNG3dOv6UuUKKFdu3ZpypQpmjBhgs6cOaMSJUrovvvu0xtvvCFfX1/FxMToySefdAi40u35iIiI0Pnz59WqVSvFxMRo7ty5unTpkkqXLq1GjRrpiy++UKlSpWznHD58WP7+/nb9eHh46Ndff1WFChUUGBio1157zfYPc2T+eejQoXc9TgAAYA4nKy/P2wIMlJqaevspC0Pel7NH0dxPAAAAeXJ6Svt86zvz8zslJUU+Pj7Ztiu0a3IBAACA7BByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYx7WgCwAK2qHX2sjHx6egywAAAH8g7uQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIzjWtAFAAWt1tgNcvYoWqA1nJ7SvkCvDwCAabiTCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGCcQhFyLctSq1at1KZNG4djc+bMka+vr86ePav169erWbNm8vb2VtGiRVW/fn0tXrzYrn1sbKycnJyUnJzs0NcDDzygqKgo258DAwPl5OSkXbt22bUbMmSImjVrZrcvNTVVr776qmrWrKkiRYqoVKlSql+/vl5//XVdvnw527HFxcWpTp068vLyUqdOneza3rp1Sw8++KD27t2b/YvzGzt27NBjjz2mEiVKyNPTU/fdd5+mT5+u9PR0u3ZOTk4O28MPP5xj30lJSXrhhRdUtWpVeXh4qGLFiurYsaO++OILh7aTJk2Si4uLpkyZ4nAsPT1dkydPVmhoqIoUKaKSJUuqYcOGWrRoka1Nnz59sqyxbdu2dq9bhw4dVLZsWXl6eiowMFARERH66aef8vRaAQAAsxWKkOvk5KRFixZp9+7dmj9/vm1/QkKCXn75Zb355pv66KOP9Pjjj6tx48bavXu3vvvuO0VGRmrAgAEaMWLEXV/b09NTL7/8co5tfv75Z1tQGzFihHbv3q2vv/5aY8eO1f79+7V8+fJsz+3fv79atGihffv2KTk5WZMmTbIdmzZtmh5++GHVr18/1zrXrl2rRx99VBUqVNDWrVt19OhRvfTSS5o4caIiIyNlWZZd+0WLFikxMdG2ffzxx9n2ffr0adWtW1dbtmzR66+/roMHD+rzzz9X8+bNNWjQIIf2ixYt0siRIxUTE+NwLCoqSrNmzdL48eN15MgRbd26Vc8884zDDwJt27a1qy8xMVErVqyQJF24cEGtWrVS6dKltWHDBsXHxysmJkb+/v66du1arq8VAAAwn2tBF5BXFStW1JtvvqnBgwcrPDxcgYGB6tevn1q2bKmWLVuqWrVqGjJkiF1IHD58uNzd3fXiiy+qa9euatCgwR1f97nnntPcuXP16aef6rHHHsuyzejRo3X27FkdO3ZM5cuXt+0PDQ1Vhw4dHALmb8XHx2vZsmUKDg5W9+7dtX79eknSqVOnFBMTo2+//TbXGq9evapnnnlGnTp10oIFC2z7+/fvr3LlyqlTp056//33FRERYTtWvHhx+fn55dq3JA0cOFBOTk7as2ePihUrZttfs2ZN/fOf/7Rru23bNl2/fl3jxo3T0qVL9eWXX6pp06a24+vWrdPAgQPVtWtX277atWs7XNPDwyPb+nbs2KHU1FQtXLhQrq63/wpXqVJFLVq0yNN4AACA+QrFndxMvXv3VsuWLdW3b1+99dZbOnTokBYsWKAPPvhAaWlpWd6xfe655+Tl5WW7C3inAgMDNWDAAI0aNUoZGRkOxzMyMrRq1Sr17NnTLuD+lpOTU7b9165dW5s2bdKtW7f0xRdf6P7775ckDRgwQK+//rq8vb1zrXHjxo26dOlSluPv2LGjgoOD73r8P//8sz7//HMNGjTILuBmKl68uN2fo6Oj1b17d7m5ual79+6Kjo62O+7n56ctW7bo4sWLd1VPZh+3bt3S2rVrc/wB4vdu3Lih1NRUuw0AAJipUIVcSVqwYIGOHDmiIUOGaP78+SpbtqyOHz8uX19f+fv7O7R3d3dX1apVdfz48bu+5r///W8lJCRo2bJlDscuXryo5ORkhYSE2O2vW7euvLy85OXlpe7du2fb98KFC/XBBx+oWrVqcnd316hRo7R06VLbmuI2bdqoevXq+ve//51tH5ljq1GjRpbHQ0NDHcbfvXt3W31eXl768MMPszz3+++/l2VZCg0Nzfb6mVJTU7V69Wr17NlTktSzZ0998MEHdmFyxowZunjxovz8/HT//fdrwIAB+uyzzxz6Wr9+vV19Xl5eGj9+vCSpYcOGGj16tHr06KHSpUurXbt2euONN3T+/Pkc65s8ebJ8fX1tW8WKFXMdEwAAKJwKXcgtW7asnn32WdWoUUNPPvlkns6xLCvHu6m5KVOmjEaMGKH//Oc/unnzZpZtft//2rVrtX//frVp00bXr1/Ptu+aNWtq27ZtOnPmjJYvX660tDRFRUXprbfe0gsvvKAmTZrowIEDWrNmjdatW5djndnd1cxq/DNnztT+/fttW+vWrXPsMy+v3/Lly1W1alXb8oMHHnhAVatW1cqVK21twsLCdOjQIe3atUt9+/bV+fPn1bFjR/Xv39+ur+bNm9vVt3//frv1vxMnTlRSUpLmzZunsLAwzZs3T6GhoTp48GC29Y0aNUopKSm27dy5c7mOCQAAFE6FLuRKkqurq20tpiQFBwcrJSVFP/74o0Pbmzdv6tSpUwoKCpIk+fj4SJJSUlIc2iYnJ8vX1zfLaw4bNkzXr1/XnDlz7PaXKVNGxYsX19GjR+32V6pUSdWrV8/TcoPfGjp0qIYMGaIKFSooNjZWXbp0UbFixdS+fXvFxsZmeU5wcLCk2+t7s3L06FHb+DP5+fmpevXqti2rpQiSFBQUJCcnp2z7/q2YmBgdPnzYNj+urq46fPiww5IFZ2dn1a9fX0OHDtXatWu1ePFiRUdHKyEhwdamWLFidvVVr15dJUuWtOunVKlS6tq1q6ZPn674+HgFBARo2rRp2dbn4eEhHx8fuw0AAJipUIbc3+vcubNcXV01ffp0h2Pz5s3T1atXbUsGgoKC5Ozs7PBYrsTERP3www8Oyw4yeXl56dVXX9XEiRPtfv3u7Oysbt266b333tMPP/xwT+P44osvdPToUQ0ePFjS7cdtpaWlSZLS0tIcHgWWKTw8XCVLlsxy/B9//LFOnDiR45KJnJQsWVJt2rTR22+/ratXrzocz3wU28GDB/XNN98oNjbW7u7rl19+qb179+rQoUPZXiMsLEySsuw/r9zd3VWtWrV76gMAAJjDiJBbqVIlvf7665o1a5bGjBmjo0eP6uTJk5oxY4ZGjhyp4cOH256s4O3treeee07Dhw/Xhx9+qISEBH399dfq3r27atSoofDw8Gyv8+yzz8rX19fhS1yTJk1S+fLl1aBBA8XExOi7777TyZMntXbtWu3cuVMuLi65juH69esaNGiQFixYIGfn29PSpEkTvf322zpw4IBWr16tJk2aZHlusWLFNH/+fH300Ud69tln9d133+n06dOKjo5Wnz591KVLF3Xr1i2vL6eDOXPmKD09XQ899JBWr16tEydOKD4+XrNnz1ajRo0k3f7C2UMPPaSmTZuqVq1atu3hhx9Wo0aNbHdzu3TpopkzZ2r37t06c+aMYmNjNWjQIAUHB9ut+71x44aSkpLstsxn4K5fv149e/bU+vXrdfz4cR07dkzTpk3Tp59+qscff/yuxwkAAMxRaB4hlpuhQ4eqWrVqmjZtmt58802lp6erZs2amjt3rvr27WvXdubMmfL399fo0aN1+vRplS1bVs2bN9fKlSvtlkH8npubm8aPH68ePXrY7S9VqpT27NmjqVOn6o033lBCQoKcnZ0VFBSkiIgIDRkyJNf6x40bpw4dOuiBBx6w7Zs9e7Z69Oihpk2bqkePHurcuXO253fp0kVbt27VpEmT1LRpU12/fl3Vq1fXmDFjNGTIkHtak1ylShXt27dPEydO1PDhw5WYmKgyZcqobt26mjt3rm7evKn33nsv2+cJd+7cWZMnT9bUqVPVpk0brVixQpMnT1ZKSor8/PzUokULRUVF2b32n3/+ucMXCUNCQnT06FGFhYWpaNGiGj58uM6dOycPDw8FBQVp4cKF6tWr112PEwAAmMPJupNnMAEGSU1Nvf2UhSHvy9mjaIHWcnpK+wK9PgAAhUXm53dKSkqO368xYrkCAAAA8FuEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwjmtBFwAUtEOvtZGPj09BlwEAAP5A3MkFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYh5ALAAAA4xByAQAAYBxCLgAAAIxDyAUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOMQcgEAAGAcQi4AAACMQ8gFAACAcQi5AAAAMA4hFwAAAMYh5AIAAMA4hFwAAAAYx7WgCwAKimVZkqTU1NQCrgQAAORV5ud25ud4dgi5+Nu6dOmSJKlixYoFXAkAALhTV65cka+vb7bHCbn42ypZsqQk6ezZszm+SQqr1NRUVaxYUefOnZOPj09Bl5MvTB+j6eOTGKMJTB+fZP4YC9v4LMvSlStXFBAQkGM7Qi7+tpydby9J9/X1LRRv6rvl4+Nj9Pgk88do+vgkxmgC08cnmT/GwjS+vNyc4otnAAAAMA4hFwAAAMYh5OJvy8PDQ2PHjpWHh0dBl5IvTB+fZP4YTR+fxBhNYPr4JPPHaOr4nKzcnr8AAAAAFDLcyQUAAIBxCLkAAAAwDiEXAAAAxiHkAgAAwDiEXBhtzpw5qlKlijw9PVW3bl199dVXObbftm2b6tatK09PT1WtWlXz5s37kyq9M5MnT1b9+vXl7e2tsmXL6oknntCxY8dyPCc2NlZOTk4O29GjR/+kqu9MVFSUQ61+fn45nlNY5k+SAgMDs5yPQYMGZdm+MMzfl19+qY4dOyogIEBOTk768MMP7Y5blqWoqCgFBASoSJEiatasmQ4fPpxrv6tXr1ZYWJg8PDwUFhamtWvX5tMIcpfTGNPS0vTyyy/rvvvuU7FixRQQEKCnn35aP/74Y459Ll68OMu5/fXXX/N5NI5ym8M+ffo41NmwYcNc+y0scygpy7lwcnLSG2+8kW2ff6U5zMvngwnvxbwg5MJYq1at0pAhQzRmzBjFxcXpkUceUbt27XT27Nks2yckJOixxx7TI488ori4OI0ePVovvviiVq9e/SdXnrtt27Zp0KBB2rVrlzZt2qRbt24pPDxcV69ezfXcY8eOKTEx0bYFBQX9CRXfnZo1a9rVevDgwWzbFqb5k6S9e/fajW3Tpk2SpK5du+Z43l95/q5evaratWvrrbfeyvL466+/rhkzZuitt97S3r175efnp9atW+vKlSvZ9rlz505FRESoV69eOnDggHr16qVu3bpp9+7d+TWMHOU0xmvXrmnfvn169dVXtW/fPq1Zs0bHjx9Xp06dcu3Xx8fHbl4TExPl6emZH0PIUW5zKElt27a1q/PTTz/Nsc/CNIeSHOYhJiZGTk5O6ty5c479/lXmMC+fDya8F/PEAgz10EMPWQMGDLDbFxoaar3yyitZth85cqQVGhpqt++5556zGjZsmG81/lEuXLhgSbK2bduWbZutW7dakqzLly//eYXdg7Fjx1q1a9fOc/vCPH+WZVkvvfSSVa1aNSsjIyPL44Vt/iRZa9eutf05IyPD8vPzs6ZMmWLb9+uvv1q+vr7WvHnzsu2nW7duVtu2be32tWnTxoqMjPzDa75Tvx9jVvbs2WNJss6cOZNtm0WLFlm+vr5/bHF/gKzG17t3b+vxxx+/o34K+xw+/vjjVosWLXJs81edQ8ty/Hww8b2YHe7kwkg3b97Ut99+q/DwcLv94eHh2rFjR5bn7Ny506F9mzZt9M033ygtLS3fav0jpKSkSJJKliyZa9s6derI399fLVu21NatW/O7tHty4sQJBQQEqEqVKoqMjNSpU6eybVuY5+/mzZt677339M9//lNOTk45ti1M8/dbCQkJSkpKspsjDw8PPfroo9m+J6Xs5zWnc/5KUlJS5OTkpOLFi+fY7pdfflHlypVVoUIFdejQQXFxcX9OgXchNjZWZcuWVXBwsJ555hlduHAhx/aFeQ7Pnz+vTz75RP369cu17V91Dn//+fB3ei8ScmGkn376Senp6SpXrpzd/nLlyikpKSnLc5KSkrJsf+vWLf3000/5Vuu9sixLw4YN08MPP6xatWpl287f318LFizQ6tWrtWbNGoWEhKhly5b68ssv/8Rq865BgwZaunSpNmzYoHfeeUdJSUlq3LixLl26lGX7wjp/kvThhx8qOTlZffr0ybZNYZu/38t8393JezLzvDs956/i119/1SuvvKIePXrIx8cn23ahoaFavHixPv74Y61YsUKenp5q0qSJTpw48SdWmzft2rXTsmXLtGXLFk2fPl179+5VixYtdOPGjWzPKcxzuGTJEnl7e+sf//hHju3+qnOY1efD3+m96FrQBQD56fd3xSzLyvFOWVbts9r/VzJ48GB999132r59e47tQkJCFBISYvtzo0aNdO7cOU2bNk1NmzbN7zLvWLt27Wz/fd9996lRo0aqVq2alixZomHDhmV5TmGcP0mKjo5Wu3btFBAQkG2bwjZ/2bnT9+TdnlPQ0tLSFBkZqYyMDM2ZMyfHtg0bNrT78laTJk304IMP6v/9v/+n2bNn53epdyQiIsL237Vq1VK9evVUuXJlffLJJzkGwcI4h5IUExOjp556Kte1tX/VOczp8+Hv8F7kTi6MVLp0abm4uDj8hHnhwgWHn0Qz+fn5Zdne1dVVpUqVyrda78ULL7ygjz/+WFu3blWFChXu+PyGDRsW+J2GvCpWrJjuu+++bOstjPMnSWfOnNHmzZvVv3//Oz63MM1f5pMx7uQ9mXnenZ5T0NLS0tStWzclJCRo06ZNOd7FzYqzs7Pq169fKObW399flStXzrHWwjiHkvTVV1/p2LFjd/Xe/CvMYXafD3+n9yIhF0Zyd3dX3bp1bd9Yz7Rp0yY1btw4y3MaNWrk0H7jxo2qV6+e3Nzc8q3Wu2FZlgYPHqw1a9Zoy5YtqlKlyl31ExcXJ39//z+4uvxx48YNxcfHZ1tvYZq/31q0aJHKli2r9u3b3/G5hWn+qlSpIj8/P7s5unnzprZt25bte1LKfl5zOqcgZQbcEydOaPPmzXf1A5ZlWdq/f3+hmNtLly7p3LlzOdZa2OYwU3R0tOrWravatWvf8bkFOYe5fT78Xd6Lkni6Asy1cuVKy83NzYqOjraOHDliDRkyxCpWrJh1+vRpy7Is65VXXrF69epla3/q1CmraNGi1tChQ60jR45Y0dHRlpubm/XBBx8U1BCy9fzzz1u+vr5WbGyslZiYaNuuXbtma/P78c2cOdNau3atdfz4cevQoUPWK6+8YkmyVq9eXRBDyNXw4cOt2NhY69SpU9auXbusDh06WN7e3kbMX6b09HSrUqVK1ssvv+xwrDDO35UrV6y4uDgrLi7OkmTNmDHDiouLsz1ZYMqUKZavr6+1Zs0a6+DBg1b37t0tf39/KzU11dZHr1697J6A8vXXX1suLi7WlClTrPj4eGvKlCmWq6urtWvXrj99fJaV8xjT0tKsTp06WRUqVLD2799v9968ceOGrY/fjzEqKsr6/PPPrZMnT1pxcXFW3759LVdXV2v37t1/qfFduXLFGj58uLVjxw4rISHB2rp1q9WoUSOrfPnyxsxhppSUFKto0aLW3Llzs+zjrzyHefl8MOG9mBeEXBjt7bfftipXrmy5u7tbDz74oN0jtnr37m09+uijdu1jY2OtOnXqWO7u7lZgYGC2/wdX0CRluS1atMjW5vfjmzp1qlWtWjXL09PTKlGihPXwww9bn3zyyZ9ffB5FRERY/v7+lpubmxUQEGD94x//sA4fPmw7XpjnL9OGDRssSdaxY8ccjhXG+ct8zNnvt969e1uWdfvRRWPHjrX8/PwsDw8Pq2nTptbBgwft+nj00Udt7TP997//tUJCQiw3NzcrNDS0QIN9TmNMSEjI9r25detWWx+/H+OQIUOsSpUqWe7u7laZMmWs8PBwa8eOHX/+4Kycx3ft2jUrPDzcKlOmjOXm5mZVqlTJ6t27t3X27Fm7PgrzHGaaP3++VaRIESs5OTnLPv7Kc5iXzwcT3ot54WRZ//83MwAAAABDsCYXAAAAxiHkAgAAwDiEXAAAABiHkAsAAADjEHIBAABgHEIuAAAAjEPIBQAAgHEIuQAAADAOIRcAAADGIeQCAADAOIRcAAAAGIeQCwAAAOP8f1OBAo/IpC4wAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "age_bkt_df.select(\n",
    "    (F.as_double((F.sum(\"YOUNG_\") ) * 100 / F.count(\"YOUNG_\"))).as_(\"YOUNG % OF CASES\"),\n",
    "    (F.as_double((F.sum(\"ADULT_\")) * 100 / F.count(\"ADULT_\"))).as_(\"ADULT % OF CASES\"),\n",
    "    (F.as_double((F.sum(\"OLD_ADULT_\")) * 100 / F.count(\"OLD_ADULT_\"))).as_(\"OLD_ADULT % OF CASES\")\n",
    ").to_pandas().T\\\n",
    ".plot.barh()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "52a6b530-21e8-45ca-b522-e4843799a203",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-----------------------------\n",
      "|\"IS_USMER_NULL\"  |\"COUNT\"  |\n",
      "-----------------------------\n",
      "|NO               |1048575  |\n",
      "-----------------------------\n",
      "\n",
      "------------------------------------\n",
      "|\"IS_MEDICAL_UNIT_NULL\"  |\"COUNT\"  |\n",
      "------------------------------------\n",
      "|NO                      |1048575  |\n",
      "------------------------------------\n",
      "\n",
      "---------------------------\n",
      "|\"IS_SEX_NULL\"  |\"COUNT\"  |\n",
      "---------------------------\n",
      "|NO             |1048575  |\n",
      "---------------------------\n",
      "\n",
      "------------------------------------\n",
      "|\"IS_PATIENT_TYPE_NULL\"  |\"COUNT\"  |\n",
      "------------------------------------\n",
      "|NO                      |1048575  |\n",
      "------------------------------------\n",
      "\n",
      "---------------------------------\n",
      "|\"IS_DATE_DIED_NULL\"  |\"COUNT\"  |\n",
      "---------------------------------\n",
      "|NO                   |1048575  |\n",
      "---------------------------------\n",
      "\n",
      "-------------------------------\n",
      "|\"IS_INTUBED_NULL\"  |\"COUNT\"  |\n",
      "-------------------------------\n",
      "|YES                |855869   |\n",
      "|NO                 |192706   |\n",
      "-------------------------------\n",
      "\n",
      "---------------------------------\n",
      "|\"IS_PNEUMONIA_NULL\"  |\"COUNT\"  |\n",
      "---------------------------------\n",
      "|NO                   |1032572  |\n",
      "|YES                  |16003    |\n",
      "---------------------------------\n",
      "\n",
      "---------------------------\n",
      "|\"IS_AGE_NULL\"  |\"COUNT\"  |\n",
      "---------------------------\n",
      "|NO             |1048354  |\n",
      "|YES            |221      |\n",
      "---------------------------\n",
      "\n",
      "--------------------------------\n",
      "|\"IS_PREGNANT_NULL\"  |\"COUNT\"  |\n",
      "--------------------------------\n",
      "|NO                  |525064   |\n",
      "|YES                 |523511   |\n",
      "--------------------------------\n",
      "\n",
      "--------------------------------\n",
      "|\"IS_DIABETES_NULL\"  |\"COUNT\"  |\n",
      "--------------------------------\n",
      "|NO                  |1048575  |\n",
      "--------------------------------\n",
      "\n",
      "----------------------------\n",
      "|\"IS_COPD_NULL\"  |\"COUNT\"  |\n",
      "----------------------------\n",
      "|NO              |1048575  |\n",
      "----------------------------\n",
      "\n",
      "------------------------------\n",
      "|\"IS_ASTHMA_NULL\"  |\"COUNT\"  |\n",
      "------------------------------\n",
      "|NO                |1048575  |\n",
      "------------------------------\n",
      "\n",
      "-------------------------------\n",
      "|\"IS_INMSUPR_NULL\"  |\"COUNT\"  |\n",
      "-------------------------------\n",
      "|NO                 |1048575  |\n",
      "-------------------------------\n",
      "\n",
      "------------------------------------\n",
      "|\"IS_HIPERTENSION_NULL\"  |\"COUNT\"  |\n",
      "------------------------------------\n",
      "|NO                      |1048575  |\n",
      "------------------------------------\n",
      "\n",
      "-------------------------------------\n",
      "|\"IS_OTHER_DISEASE_NULL\"  |\"COUNT\"  |\n",
      "-------------------------------------\n",
      "|NO                       |1048575  |\n",
      "-------------------------------------\n",
      "\n",
      "--------------------------------------\n",
      "|\"IS_CARDIOVASCULAR_NULL\"  |\"COUNT\"  |\n",
      "--------------------------------------\n",
      "|NO                        |1048575  |\n",
      "--------------------------------------\n",
      "\n",
      "-------------------------------\n",
      "|\"IS_OBESITY_NULL\"  |\"COUNT\"  |\n",
      "-------------------------------\n",
      "|NO                 |1048575  |\n",
      "-------------------------------\n",
      "\n",
      "-------------------------------------\n",
      "|\"IS_RENAL_CHRONIC_NULL\"  |\"COUNT\"  |\n",
      "-------------------------------------\n",
      "|NO                       |1048575  |\n",
      "-------------------------------------\n",
      "\n",
      "-------------------------------\n",
      "|\"IS_TOBACCO_NULL\"  |\"COUNT\"  |\n",
      "-------------------------------\n",
      "|NO                 |1048575  |\n",
      "-------------------------------\n",
      "\n",
      "--------------------------------------------\n",
      "|\"IS_CLASIFFICATION_FINAL_NULL\"  |\"COUNT\"  |\n",
      "--------------------------------------------\n",
      "|NO                              |1048575  |\n",
      "--------------------------------------------\n",
      "\n",
      "---------------------------\n",
      "|\"IS_ICU_NULL\"  |\"COUNT\"  |\n",
      "---------------------------\n",
      "|YES            |856032   |\n",
      "|NO             |192543   |\n",
      "---------------------------\n",
      "\n",
      "------------------------------\n",
      "|\"IS_TARGET_NULL\"  |\"COUNT\"  |\n",
      "------------------------------\n",
      "|NO                |1048575  |\n",
      "------------------------------\n",
      "\n",
      "-------------------------------\n",
      "|\"IS_AGE_BKT_NULL\"  |\"COUNT\"  |\n",
      "-------------------------------\n",
      "|NO                 |1048575  |\n",
      "-------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for col in snowpark_df.columns:\n",
    "    snowpark_df.with_column(f\"IS_{col}_NULL\",F.when(F.is_null(col),\"YES\").otherwise(\"NO\")).group_by(f'IS_{col}_NULL').count().show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "48763151-89c8-443f-aa9a-fcea1ac53dd4",
   "metadata": {},
   "outputs": [],
   "source": [
    "snowpark_df = snowpark_df.drop([\"ICU\",\"DATE_DIED\",\"CLASIFFICATION_FINAL\",\"AGE_BKT\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "5a2991e7-b654-42b6-b721-899906722a5c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------\n",
      "|\"TARGET\"  |\"COUNT\"  |\n",
      "----------------------\n",
      "|1         |41926    |\n",
      "|0         |36262    |\n",
      "----------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "snowpark_df.dropna().group_by(\"TARGET\").count().show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "0aa788d8-54ce-4e0d-ab94-6c8496d297f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "snowpark_df = snowpark_df.dropna()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8994641f-04b9-4db7-bc10-8fd50f49d3eb",
   "metadata": {},
   "source": [
    "Also you can check the queries generated as below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "af153fdd-65c5-40ae-91e0-c4442fb3cbf6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# snowpark_df.queries['queries'][0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "87e5b63c-5acb-474d-ad4d-c613a52057bc",
   "metadata": {},
   "source": [
    "``` sql \n",
    "SELECT\n",
    "    *\n",
    "FROM\n",
    "    (\n",
    "        SELECT\n",
    "            \"USMER\",\n",
    "            \"MEDICAL_UNIT\",\n",
    "            \"SEX\",\n",
    "            \"PATIENT_TYPE\",\n",
    "            \"DATE_DIED\",\n",
    "            \"INTUBED\",\n",
    "            \"PNEUMONIA\",\n",
    "            \"AGE\",\n",
    "            \"PREGNANT\",\n",
    "            \"DIABETES\",\n",
    "            \"COPD\",\n",
    "            \"ASTHMA\",\n",
    "            \"INMSUPR\",\n",
    "            \"HIPERTENSION\",\n",
    "            \"OTHER_DISEASE\",\n",
    "            \"CARDIOVASCULAR\",\n",
    "            \"OBESITY\",\n",
    "            \"RENAL_CHRONIC\",\n",
    "            \"TOBACCO\",\n",
    "            \"CLASIFFICATION_FINAL\",\n",
    "            \"TARGET\"\n",
    "        FROM\n",
    "            (\n",
    "                SELECT\n",
    "                    CASE\n",
    "                        WHEN (\"USMER\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"USMER\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"USMER\"\n",
    "                    END AS \"USMER\",\n",
    "                    CASE\n",
    "                        WHEN (\"MEDICAL_UNIT\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"MEDICAL_UNIT\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"MEDICAL_UNIT\"\n",
    "                    END AS \"MEDICAL_UNIT\",\n",
    "                    CASE\n",
    "                        WHEN (\"SEX\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"SEX\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"SEX\"\n",
    "                    END AS \"SEX\",\n",
    "                    CASE\n",
    "                        WHEN (\"PATIENT_TYPE\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"PATIENT_TYPE\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"PATIENT_TYPE\"\n",
    "                    END AS \"PATIENT_TYPE\",\n",
    "                    \"DATE_DIED\",\n",
    "                    CASE\n",
    "                        WHEN (\"INTUBED\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"INTUBED\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"INTUBED\"\n",
    "                    END AS \"INTUBED\",\n",
    "                    CASE\n",
    "                        WHEN (\"PNEUMONIA\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"PNEUMONIA\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"PNEUMONIA\"\n",
    "                    END AS \"PNEUMONIA\",\n",
    "                    CASE\n",
    "                        WHEN (\"AGE\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"AGE\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"AGE\"\n",
    "                    END AS \"AGE\",\n",
    "                    CASE\n",
    "                        WHEN (\"PREGNANT\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"PREGNANT\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"PREGNANT\"\n",
    "                    END AS \"PREGNANT\",\n",
    "                    CASE\n",
    "                        WHEN (\"DIABETES\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"DIABETES\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"DIABETES\"\n",
    "                    END AS \"DIABETES\",\n",
    "                    CASE\n",
    "                        WHEN (\"COPD\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"COPD\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"COPD\"\n",
    "                    END AS \"COPD\",\n",
    "                    CASE\n",
    "                        WHEN (\"ASTHMA\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"ASTHMA\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"ASTHMA\"\n",
    "                    END AS \"ASTHMA\",\n",
    "                    CASE\n",
    "                        WHEN (\"INMSUPR\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"INMSUPR\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"INMSUPR\"\n",
    "                    END AS \"INMSUPR\",\n",
    "                    CASE\n",
    "                        WHEN (\"HIPERTENSION\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"HIPERTENSION\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"HIPERTENSION\"\n",
    "                    END AS \"HIPERTENSION\",\n",
    "                    CASE\n",
    "                        WHEN (\"OTHER_DISEASE\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"OTHER_DISEASE\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"OTHER_DISEASE\"\n",
    "                    END AS \"OTHER_DISEASE\",\n",
    "                    CASE\n",
    "                        WHEN (\"CARDIOVASCULAR\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"CARDIOVASCULAR\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"CARDIOVASCULAR\"\n",
    "                    END AS \"CARDIOVASCULAR\",\n",
    "                    CASE\n",
    "                        WHEN (\"OBESITY\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"OBESITY\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"OBESITY\"\n",
    "                    END AS \"OBESITY\",\n",
    "                    CASE\n",
    "                        WHEN (\"RENAL_CHRONIC\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"RENAL_CHRONIC\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"RENAL_CHRONIC\"\n",
    "                    END AS \"RENAL_CHRONIC\",\n",
    "                    CASE\n",
    "                        WHEN (\"TOBACCO\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"TOBACCO\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"TOBACCO\"\n",
    "                    END AS \"TOBACCO\",\n",
    "                    CASE\n",
    "                        WHEN (\"CLASIFFICATION_FINAL\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"CLASIFFICATION_FINAL\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"CLASIFFICATION_FINAL\"\n",
    "                    END AS \"CLASIFFICATION_FINAL\",\n",
    "                    CASE\n",
    "                        WHEN (\"ICU\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"ICU\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"ICU\"\n",
    "                    END AS \"ICU\",\n",
    "                    CASE\n",
    "                        WHEN (\"TARGET\" = 97 :: INT) THEN NULL\n",
    "                        WHEN (\"TARGET\" = 99 :: INT) THEN NULL\n",
    "                        ELSE \"TARGET\"\n",
    "                    END AS \"TARGET\"\n",
    "                FROM\n",
    "                    (\n",
    "                        SELECT\n",
    "                            \"USMER\",\n",
    "                            \"MEDICAL_UNIT\",\n",
    "                            \"SEX\",\n",
    "                            \"PATIENT_TYPE\",\n",
    "                            \"DATE_DIED\",\n",
    "                            \"INTUBED\",\n",
    "                            \"PNEUMONIA\",\n",
    "                            \"AGE\",\n",
    "                            \"PREGNANT\",\n",
    "                            \"DIABETES\",\n",
    "                            \"COPD\",\n",
    "                            \"ASTHMA\",\n",
    "                            \"INMSUPR\",\n",
    "                            \"HIPERTENSION\",\n",
    "                            \"OTHER_DISEASE\",\n",
    "                            \"CARDIOVASCULAR\",\n",
    "                            \"OBESITY\",\n",
    "                            \"RENAL_CHRONIC\",\n",
    "                            \"TOBACCO\",\n",
    "                            \"CLASIFFICATION_FINAL\",\n",
    "                            \"ICU\",\n",
    "                            CASE\n",
    "                                WHEN (\"CLASIFFICATION_FINAL\" < 4 :: INT) THEN 1 :: INT\n",
    "                                ELSE 0 :: INT\n",
    "                            END AS \"TARGET\"\n",
    "                        FROM\n",
    "                            (\n",
    "                                SELECT\n",
    "                                    *\n",
    "                                FROM\n",
    "                                    (COVID19_RECORDS)\n",
    "                            )\n",
    "                    )\n",
    "            )\n",
    "    )\n",
    "WHERE\n",
    "    (\n",
    "        (\n",
    "            (\n",
    "                (\n",
    "                    (\n",
    "                        (\n",
    "                            (\n",
    "                                (\n",
    "                                    (\n",
    "                                        (\n",
    "                                            (\n",
    "                                                (\n",
    "                                                    (\n",
    "                                                        (\n",
    "                                                            (\n",
    "                                                                (\n",
    "                                                                    (\n",
    "                                                                        (\n",
    "                                                                            (\n",
    "                                                                                (\n",
    "                                                                                    (\n",
    "                                                                                        iff(\"AGE\" IS NULL, 0, 1) + iff(\"INTUBED\" IS NULL, 0, 1)\n",
    "                                                                                    ) + iff(\"PREGNANT\" IS NULL, 0, 1)\n",
    "                                                                                ) + iff(\"HIPERTENSION\" IS NULL, 0, 1)\n",
    "                                                                            ) + iff(\"PNEUMONIA\" IS NULL, 0, 1)\n",
    "                                                                        ) + iff(\"OTHER_DISEASE\" IS NULL, 0, 1)\n",
    "                                                                    ) + iff(\"TOBACCO\" IS NULL, 0, 1)\n",
    "                                                                ) + iff(\"INMSUPR\" IS NULL, 0, 1)\n",
    "                                                            ) + iff(\"DIABETES\" IS NULL, 0, 1)\n",
    "                                                        ) + iff(\"TARGET\" IS NULL, 0, 1)\n",
    "                                                    ) + iff(\"COPD\" IS NULL, 0, 1)\n",
    "                                                ) + iff(\"CARDIOVASCULAR\" IS NULL, 0, 1)\n",
    "                                            ) + iff(\"MEDICAL_UNIT\" IS NULL, 0, 1)\n",
    "                                        ) + iff(\"PATIENT_TYPE\" IS NULL, 0, 1)\n",
    "                                    ) + iff(\"SEX\" IS NULL, 0, 1)\n",
    "                                ) + iff(\"DATE_DIED\" IS NULL, 0, 1)\n",
    "                            ) + iff(\"CLASIFFICATION_FINAL\" IS NULL, 0, 1)\n",
    "                        ) + iff(\"ASTHMA\" IS NULL, 0, 1)\n",
    "                    ) + iff(\"OBESITY\" IS NULL, 0, 1)\n",
    "                ) + iff(\"RENAL_CHRONIC\" IS NULL, 0, 1)\n",
    "            ) + iff(\"USMER\" IS NULL, 0, 1)\n",
    "        ) >= 21 :: INT\n",
    "    )\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "2f98e0e4-b75b-438e-97ba-4c8b7874c633",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>USMER</th>\n",
       "      <th>MEDICAL_UNIT</th>\n",
       "      <th>SEX</th>\n",
       "      <th>PATIENT_TYPE</th>\n",
       "      <th>INTUBED</th>\n",
       "      <th>PNEUMONIA</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PREGNANT</th>\n",
       "      <th>DIABETES</th>\n",
       "      <th>COPD</th>\n",
       "      <th>ASTHMA</th>\n",
       "      <th>INMSUPR</th>\n",
       "      <th>HIPERTENSION</th>\n",
       "      <th>OTHER_DISEASE</th>\n",
       "      <th>CARDIOVASCULAR</th>\n",
       "      <th>OBESITY</th>\n",
       "      <th>RENAL_CHRONIC</th>\n",
       "      <th>TOBACCO</th>\n",
       "      <th>TARGET</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>40</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>37</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>25</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>80</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>58</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>25</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>24</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>25</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>30</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   USMER  MEDICAL_UNIT  SEX  PATIENT_TYPE  INTUBED  PNEUMONIA  AGE  PREGNANT  \\\n",
       "0      2             1    1             2        2          1   40         2   \n",
       "1      2             1    1             2        2          2   37         2   \n",
       "2      2             1    1             2        2          2   25         2   \n",
       "3      2             1    1             2        2          1   80         2   \n",
       "4      2             1    1             2        1          1   58         2   \n",
       "5      2             1    1             2        1          1   48         2   \n",
       "6      2             1    1             2        2          2   25         2   \n",
       "7      2             1    1             2        2          2   24         2   \n",
       "8      2             1    1             2        2          2   25         2   \n",
       "9      2             1    1             2        2          2   30         2   \n",
       "\n",
       "   DIABETES  COPD  ASTHMA  INMSUPR  HIPERTENSION  OTHER_DISEASE  \\\n",
       "0         2     2       2        2             2              2   \n",
       "1         1     2       2        2             1              2   \n",
       "2         2     2       2        2             2              2   \n",
       "3         2     2       2        2             1              2   \n",
       "4         2     2       2        2             1              2   \n",
       "5         1     2       2        2             2              2   \n",
       "6         2     2       2        2             2              2   \n",
       "7         2     2       2        2             2              2   \n",
       "8         2     2       2        2             2              2   \n",
       "9         2     2       2        2             2              1   \n",
       "\n",
       "   CARDIOVASCULAR  OBESITY  RENAL_CHRONIC  TOBACCO  TARGET  \n",
       "0               2        2              2        2       1  \n",
       "1               2        1              2        2       1  \n",
       "2               2        2              2        2       1  \n",
       "3               2        2              2        2       1  \n",
       "4               1        1              2        2       0  \n",
       "5               2        2              2        2       0  \n",
       "6               2        2              2        2       0  \n",
       "7               2        2              2        2       0  \n",
       "8               2        2              2        2       0  \n",
       "9               2        2              2        2       0  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "snowpark_df.limit(10).to_pandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d43f52d4-2efb-4bfd-978e-b6bd4a458eed",
   "metadata": {},
   "source": [
    "Write back to snowflake"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "858afd54-c42e-40c7-bbb2-4f40229af726",
   "metadata": {},
   "outputs": [],
   "source": [
    "snowpark_df.write.save_as_table(\n",
    "    table_name='COVID19_RECORDS_PROCESSED',\n",
    "    mode='overwrite'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "7e5219af-b7f8-47eb-a9a8-00c1190583fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "del snowpark_df\n",
    "del df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "89c1d3d4-f162-4688-bc09-45d6597d4a6c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21443"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import gc\n",
    "\n",
    "gc.collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2944b87-4563-483b-a197-b904094dbade",
   "metadata": {},
   "source": [
    "# Model training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "81f74f7c-d6e7-48c8-9590-82e56d6df830",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = session.table(\"COVID19_RECORDS_PROCESSED\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "3f7dc70e-c89c-4a9a-96ed-6e761d4bd3eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "78188"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "673d88d1-a45c-40d1-8e49-01d112045b6a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------------------\n",
      "|\"SUM(\"\"TARGET\"\")\"  |\n",
      "---------------------\n",
      "|41926              |\n",
      "---------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "train_data.select(F.sum(F.col(\"TARGET\"))).show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d9c7ab6-bae5-4fa3-976c-b542182d144b",
   "metadata": {},
   "source": [
    "Let's start real model training part"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fcfac575-fcc3-4463-9c0f-7836b055d16d",
   "metadata": {},
   "source": [
    "First, create stage for machine learning models, called `ML_MODELS`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "ac816732-2f55-45cd-8dc1-43af058c5f50",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------------------------------\n",
      "|\"status\"                                    |\n",
      "----------------------------------------------\n",
      "|Stage area ML_MODELS successfully created.  |\n",
      "----------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "session.sql('CREATE OR REPLACE STAGE ML_MODELS').show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "433e13a6-f923-4e97-aa7b-c2017e3e1d13",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_dt_procedure(\n",
    "    session: Session,\n",
    "    training_table: str,\n",
    "    feature_cols: list,\n",
    "    target_col: str,\n",
    "    model_name: str,\n",
    ") -> T.Variant:\n",
    "    \n",
    "    \"\"\"\n",
    "    This will be our training procedure. Later we will register this as snowflake procedure.\n",
    "    \n",
    "    training_table: snowflake table name to be used for training task\n",
    "    feature_cols: list of columns to be used in training\n",
    "    target_col: target column to be used\n",
    "    model_name: model name to used for model saving purpose\n",
    "    \n",
    "    \"\"\"\n",
    "\n",
    "    #convert as pandas DF, rest of the steps similar to the local model training and saving.\n",
    "    local_training_data = session.table(training_table).to_pandas()\n",
    "\n",
    "    from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "    X = local_training_data[feature_cols]\n",
    "    y = local_training_data[target_col]\n",
    "\n",
    "    model = DecisionTreeClassifier()\n",
    "    model.fit(X, y)\n",
    "\n",
    "    feat_importance = pd.DataFrame(\n",
    "        model.feature_importances_, feature_cols, columns=[\"FeatImportance\"]\n",
    "    ).to_dict()\n",
    "\n",
    "    from joblib import dump\n",
    "\n",
    "    dump(model, \"/tmp/\" + model_name)\n",
    "    session.file.put(\n",
    "        \"/tmp/\" + model_name, \"@ML_MODELS\", auto_compress=False, overwrite=True\n",
    "    )\n",
    "    return feat_importance\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "a9cfd6ef-30dd-4745-ab81-97feae818c0a",
   "metadata": {},
   "outputs": [],
   "source": [
    "sproc_train_dt_model = session.sproc.register(\n",
    "                    func=train_dt_procedure, \n",
    "                    name='sproc_train_dt_model', \n",
    "                    is_permanent=True, \n",
    "                    replace=True, \n",
    "                    stage_location='@ML_MODELS', \n",
    "                    packages=[\n",
    "                        'snowflake-snowpark-python',\n",
    "                        'scikit-learn',\n",
    "                        'joblib']\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "3c794edb-9d51-471e-a67e-831437551665",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data_pd = train_data.to_pandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "9ec4ab23-384e-4e01-b85d-28ac44b2418f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#create train and test dataframes\n",
    "train_data_pd,test_data_pd = train_test_split(\n",
    "                                        train_data_pd,\n",
    "                                        stratify=train_data_pd['TARGET'],\n",
    "                                        test_size=0.1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "e1e90b49-469f-4f0c-920e-0179caf1fae5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<snowflake.snowpark.table.Table at 0x28a2c95f160>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#writing as tempoary tables for mode training and inferencing part\n",
    "session.write_pandas(train_data_pd, table_name='TRAIN_DATA_TMP', auto_create_table=True,table_type=\"temporary\")\n",
    "session.write_pandas(test_data_pd, table_name='TEST_DATA_TMP', auto_create_table=True,table_type=\"temporary\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "f2f93765-112d-45e3-8582-c0d2fe330cb9",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_cols = train_data.columns\n",
    "feature_cols.remove('TARGET')\n",
    "target_col = 'TARGET'\n",
    "model_name = 'decisiontree.model'\n",
    "\n",
    "model_response = sproc_train_dt_model('TRAIN_DATA_TMP', \n",
    "                                            feature_cols, \n",
    "                                            target_col,\n",
    "                                            model_name, \n",
    "                                            session=session\n",
    "                                           )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "25f58489-62f1-490b-985f-f28e7464668f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"FeatImportance\": {\n",
      "    \"AGE\": 0.4495406764847183,\n",
      "    \"ASTHMA\": 0.02734844145002929,\n",
      "    \"CARDIOVASCULAR\": 0.026163656824010745,\n",
      "    \"COPD\": 0.015610870661814366,\n",
      "    \"DIABETES\": 0.06058696387222489,\n",
      "    \"HIPERTENSION\": 0.06046161472931457,\n",
      "    \"INMSUPR\": 0.024452172596442383,\n",
      "    \"INTUBED\": 0.025061840298817285,\n",
      "    \"MEDICAL_UNIT\": 0.09498782681802134,\n",
      "    \"OBESITY\": 0.028268191523668378,\n",
      "    \"OTHER_DISEASE\": 0.025908224512469692,\n",
      "    \"PATIENT_TYPE\": 0,\n",
      "    \"PNEUMONIA\": 0.0412125714132776,\n",
      "    \"PREGNANT\": 0.01335446540097891,\n",
      "    \"RENAL_CHRONIC\": 0.018501343421274265,\n",
      "    \"SEX\": 0,\n",
      "    \"TOBACCO\": 0.027695748397739643,\n",
      "    \"USMER\": 0.06084539159519835\n",
      "  }\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "print(model_response)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "8ed3f0c6-c393-4f37-b0bd-2c09338efaa5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------------------------------------\n",
      "|\"name\"                        |\"size\"   |\"md5\"                             |\"last_modified\"                |\n",
      "-------------------------------------------------------------------------------------------------------------\n",
      "|ml_models/decisiontree.model  |2618592  |71ccb5ac5ad7739e8c495ee0a989b06a  |Mon, 26 Dec 2022 05:51:59 GMT  |\n",
      "-------------------------------------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "session.sql('LIST @ML_MODELS').show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "427d160d-b571-4215-87b2-2405ad102020",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_coefficients = pd.DataFrame(eval(model_response))\n",
    "\n",
    "feature_coefficients\\\n",
    ".sort_values(by='FeatImportance',ascending=False)\\\n",
    ".plot\\\n",
    ".bar(y='FeatImportance', figsize=(12,5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "09c16875-f96b-4752-9ea1-c47305858bdf",
   "metadata": {},
   "source": [
    "Let's create UDF for getting model predictions for unseen data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "c8f9f19e-449f-44af-916f-b842bab64dfe",
   "metadata": {},
   "outputs": [],
   "source": [
    "def udf_predict_risk_score(*args) -> float:\n",
    "    import os\n",
    "    import sys\n",
    "    from joblib import load\n",
    "    \n",
    "    IMPORT_DIRECTORY_NAME = \"snowflake_import_directory\"\n",
    "    import_dir = sys._xoptions[IMPORT_DIRECTORY_NAME]\n",
    "    model_name = 'decisiontree.model'\n",
    "    model = load(import_dir+model_name)\n",
    "    scored_data = model.predict(pd.DataFrame([args]))[0]\n",
    "    return scored_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "800f6a12-12d6-40c7-a913-8e56fbe663e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "udf_risk_score_model = session.udf.register(\n",
    "                            func=udf_predict_risk_score, \n",
    "                            name=\"udf_risk_score_model\", \n",
    "                            stage_location='@ML_MODELS',\n",
    "                            input_types=[T.FloatType()]*len(feature_cols),\n",
    "                            return_type = T.FloatType(),\n",
    "                            replace=True, \n",
    "                            is_permanent=True, \n",
    "                            imports=['@ML_MODELS/decisiontree.model'],\n",
    "                            packages=['scikit-learn==1.1.1','pandas','joblib'], \n",
    "                            session=session\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "1a97935f-ff3e-43b5-a0b8-271d4eff7e5a",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data_sf = session.table(\"TEST_DATA_TMP\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "90295167-04f6-4fd9-86a2-a149490a37fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data_sf = test_data_sf.with_column('PREDICTION', udf_risk_score_model(*feature_cols))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "e3f6932e-bf62-4377-b44d-690c5a7c9341",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SELECT \"USMER\", \"MEDICAL_UNIT\", \"SEX\", \"PATIENT_TYPE\", \"INTUBED\", \"PNEUMONIA\", \"AGE\", \"PREGNANT\", \"DIABETES\", \"COPD\", \"ASTHMA\", \"INMSUPR\", \"HIPERTENSION\", \"OTHER_DISEASE\", \"CARDIOVASCULAR\", \"OBESITY\", \"RENAL_CHRONIC\", \"TOBACCO\", \"TARGET\", udf_risk_score_model(\"USMER\", \"MEDICAL_UNIT\", \"SEX\", \"PATIENT_TYPE\", \"INTUBED\", \"PNEUMONIA\", \"AGE\", \"PREGNANT\", \"DIABETES\", \"COPD\", \"ASTHMA\", \"INMSUPR\", \"HIPERTENSION\", \"OTHER_DISEASE\", \"CARDIOVASCULAR\", \"OBESITY\", \"RENAL_CHRONIC\", \"TOBACCO\") AS \"PREDICTION\" FROM ( SELECT  *  FROM (TEST_DATA_TMP))\n"
     ]
    }
   ],
   "source": [
    "print(test_data_sf.queries['queries'][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "f18c70cb-6e66-4bf2-9d76-b533c42592e5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['USMER',\n",
       " 'MEDICAL_UNIT',\n",
       " 'SEX',\n",
       " 'PATIENT_TYPE',\n",
       " 'INTUBED',\n",
       " 'PNEUMONIA',\n",
       " 'AGE',\n",
       " 'PREGNANT',\n",
       " 'DIABETES',\n",
       " 'COPD',\n",
       " 'ASTHMA',\n",
       " 'INMSUPR',\n",
       " 'HIPERTENSION',\n",
       " 'OTHER_DISEASE',\n",
       " 'CARDIOVASCULAR',\n",
       " 'OBESITY',\n",
       " 'RENAL_CHRONIC',\n",
       " 'TOBACCO']"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "71e2cb1d-6ab9-4362-a89c-b309128aeb59",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_model_report(test_df):\n",
    "    from sklearn.metrics import confusion_matrix,precision_score,recall_score,f1_score,ConfusionMatrixDisplay\n",
    "    \n",
    "    target = test_df.select(\"TARGET\").to_pandas()\n",
    "    predictions = test_df.select(\"PREDICTION\").to_pandas()\n",
    "    \n",
    "    cm = confusion_matrix(target,predictions)\n",
    "    precision = precision_score(target,predictions)\n",
    "    recall = recall_score(target,predictions)\n",
    "    f1= f1_score(target,predictions)\n",
    "    \n",
    "    print(f\"Model perfomances Recall: {recall:.2}, Precision: {precision:.2}\\n\")\n",
    "    \n",
    "    cmd = ConfusionMatrixDisplay(cm)\n",
    "    cmd.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "8fd24e62-b598-4a7e-b49c-0cbbd921db92",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model perfomances Recall: 0.53, Precision: 0.54\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "get_model_report(test_df=test_data_sf)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4b688fb-ee87-4304-bdc4-5a93c09c4f3e",
   "metadata": {},
   "source": [
    "Additionaly we can host locally trained model in snowflake"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "d0850d1f-b62e-4e79-a22f-074f7caafb73",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_model_locally(train):\n",
    "    from sklearn.tree import DecisionTreeClassifier\n",
    "    \n",
    "    #convert into pd dataframes\n",
    "    \n",
    "    train = train.to_pandas()\n",
    "    \n",
    "    xtrain,ytrain = train.drop('TARGET',axis=1),train['TARGET']\n",
    "    \n",
    "    model = DecisionTreeClassifier()\n",
    "    model.fit(xtrain,ytrain)\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "d4024f29-4878-4459-8998-6beb619cfc02",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data_sf = session.table(\"TRAIN_DATA_TMP\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "4b82d47a-d2c8-4bf7-a15f-d424c44d8aa0",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = train_model_locally(train_data_sf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "943de18c-bee3-44f5-8927-bfbff88da965",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['predict_risk_score.joblib']"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import joblib\n",
    "\n",
    "joblib.dump(model, 'predict_risk_score.joblib')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "a2277611-b2fc-4b67-bbfe-27d722411182",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[PutResult(source='predict_risk_score.joblib', target='predict_risk_score.joblib', source_size=2615954, target_size=2615968, source_compression='NONE', target_compression='NONE', status='UPLOADED', message='')]"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "session.file.put(\n",
    "    \"predict_risk_score.joblib\", \"@ML_MODELS\", auto_compress=False, overwrite=True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "932d01b5-3a83-4dd4-b93d-a58db0caf09a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[1;31mSignature:\u001b[0m \u001b[0msession\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0madd_import\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpath\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mstr\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mimport_path\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mUnion\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mstr\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mNoneType\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m->\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
       "\u001b[1;31mDocstring:\u001b[0m\n",
       "Registers a remote file in stage or a local file as an import of a user-defined function\n",
       "(UDF). The local file can be a compressed file (e.g., zip), a Python file (.py),\n",
       "a directory, or any other file resource. You can also find examples in\n",
       ":class:`~snowflake.snowpark.udf.UDFRegistration`.\n",
       "\n",
       "Args:\n",
       "    path: The path of a local file or a remote file in the stage. In each case:\n",
       "\n",
       "        * if the path points to a local file, this file will be uploaded to the\n",
       "          stage where the UDF is registered and Snowflake will import the file when\n",
       "          executing that UDF.\n",
       "\n",
       "        * if the path points to a local directory, the directory will be compressed\n",
       "          as a zip file and will be uploaded to the stage where the UDF is registered\n",
       "          and Snowflake will import the file when executing that UDF.\n",
       "\n",
       "        * if the path points to a file in a stage, the file will be included in the\n",
       "          imports when executing a UDF.\n",
       "\n",
       "    import_path: The relative Python import path for a UDF.\n",
       "        If it is not provided or it is None, the UDF will import the package\n",
       "        directly without any leading package/module. This argument will become\n",
       "        a no-op if the path  points to a stage file or a non-Python local file.\n",
       "\n",
       "Example::\n",
       "\n",
       "    >>> from snowflake.snowpark.types import IntegerType\n",
       "    >>> from resources.test_udf_dir.test_udf_file import mod5\n",
       "    >>> session.add_import(\"tests/resources/test_udf_dir/test_udf_file.py\", import_path=\"resources.test_udf_dir.test_udf_file\")\n",
       "    >>> mod5_and_plus1_udf = session.udf.register(\n",
       "    ...     lambda x: mod5(x) + 1,\n",
       "    ...     return_type=IntegerType(),\n",
       "    ...     input_types=[IntegerType()]\n",
       "    ... )\n",
       "    >>> session.range(1, 8, 2).select(mod5_and_plus1_udf(\"id\")).to_df(\"col1\").collect()\n",
       "    [Row(COL1=2), Row(COL1=4), Row(COL1=1), Row(COL1=3)]\n",
       "    >>> session.clear_imports()\n",
       "\n",
       "Note:\n",
       "    1. In favor of the lazy execution, the file will not be uploaded to the stage\n",
       "    immediately, and it will be uploaded when a UDF is created.\n",
       "\n",
       "    2. The Snowpark library calculates a sha256 checksum for every file/directory.\n",
       "    Each file is uploaded to a subdirectory named after the checksum for the\n",
       "    file in the stage. If there is an existing file or directory, the Snowpark\n",
       "    library will compare their checksums to determine whether it should be re-uploaded.\n",
       "    Therefore, after uploading a local file to the stage, if the user makes\n",
       "    some changes to this file and intends to upload it again, just call this\n",
       "    function with the file path again, the existing file in the stage will be\n",
       "    overwritten by the re-uploaded file.\n",
       "\n",
       "    3. Adding two files with the same file name is not allowed, because UDFs\n",
       "    can't be created with two imports with the same name.\n",
       "\n",
       "    4. This method will register the file for all UDFs created later in the current\n",
       "    session. If you only want to import a file for a specific UDF, you can use\n",
       "    ``imports`` argument in :func:`functions.udf` or\n",
       "    :meth:`session.udf.register() <snowflake.snowpark.udf.UDFRegistration.register>`.\n",
       "\u001b[1;31mFile:\u001b[0m      c:\\users\\ransaka_09914\\anaconda3\\envs\\getting_started_snowpark_python\\lib\\site-packages\\snowflake\\snowpark\\session.py\n",
       "\u001b[1;31mType:\u001b[0m      method\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "?session.add_import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "4654679d-e4a5-41d0-bdd6-dffbd69aa117",
   "metadata": {},
   "outputs": [],
   "source": [
    "session.clear_imports()\n",
    "session.clear_packages()\n",
    "session.add_import(\"@ML_MODELS/predict_risk_score.joblib\")\n",
    "session.add_packages(\"joblib==1.1.0\", \"scikit-learn==1.1.1\", \"pandas==1.3.2\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "41c4c563-6335-45fd-b63d-f0008ef8163f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from snowflake.snowpark.types import PandasSeries, PandasDataFrame\n",
    "\n",
    "\n",
    "def read_file(filename):\n",
    "    import joblib\n",
    "    import sys\n",
    "    import os\n",
    "    \n",
    "    import_dir = sys._xoptions.get(\"snowflake_import_directory\")\n",
    "    if import_dir:\n",
    "        with open(os.path.join(import_dir, filename), 'rb') as file:\n",
    "            m = joblib.load(file)\n",
    "            return m\n",
    "\n",
    "@F.udf(name = 'predict_risk_score', is_permanent = True, replace = True, stage_location = '@ML_MODELS')\n",
    "def predict_risk_score(ds: PandasSeries[dict]) -> PandasSeries[float]:\n",
    "    df = pd.io.json.json_normalize(ds)[feature_cols]\n",
    "    pipeline = read_file('predict_risk_score.joblib')\n",
    "    return pipeline.predict_proba(df)[:,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "b26b67f0-2a93-471f-ba2d-b9ada84ae6eb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "------------------------------------------------------------------------\n",
      "|\"property\"       |\"value\"                                             |\n",
      "------------------------------------------------------------------------\n",
      "|signature        |()                                                  |\n",
      "|returns          |FLOAT                                               |\n",
      "|language         |PYTHON                                              |\n",
      "|null handling    |CALLED ON NULL INPUT                                |\n",
      "|volatility       |VOLATILE                                            |\n",
      "|body             |                                                    |\n",
      "|                 |import pickle                                       |\n",
      "|                 |                                                    |\n",
      "|                 |func = pickle.loads(bytes.fromhex('800595050400...  |\n",
      "|                 |# The following comment contains the UDF source...  |\n",
      "|                 |# import pandas as pd                               |\n",
      "|                 |# def read_file(filename):                          |\n",
      "|                 |#     import joblib                                 |\n",
      "|                 |#     import sys                                    |\n",
      "|                 |#     import os                                     |\n",
      "|                 |#                                                   |\n",
      "|                 |#     import_dir = sys._xoptions.get(\"snowflake...  |\n",
      "|                 |#     if import_dir:                                |\n",
      "|                 |#         with open(os.path.join(import_dir, fi...  |\n",
      "|                 |#             m = joblib.load(file)                 |\n",
      "|                 |#             return m                              |\n",
      "|                 |# @F.udf(name = 'predict_risk_score', is_perman...  |\n",
      "|                 |# def predict_risk_score(*args) -> PandasSeries...  |\n",
      "|                 |#     df = pd.DataFrame([args])                     |\n",
      "|                 |#     pipeline = read_file('predict_risk_score....  |\n",
      "|                 |#     return pipeline.predict_proba(df)[:,1]        |\n",
      "|                 |#                                                   |\n",
      "|                 |# func = predict_risk_score                         |\n",
      "|                 |                                                    |\n",
      "|                 |                                                    |\n",
      "|                 |                                                    |\n",
      "|                 |from threading import RLock                         |\n",
      "|                 |                                                    |\n",
      "|                 |lock = RLock()                                      |\n",
      "|                 |                                                    |\n",
      "|                 |class InvokedFlag:                                  |\n",
      "|                 |    def __init__(self):                             |\n",
      "|                 |        self.invoked = False                        |\n",
      "|                 |                                                    |\n",
      "|                 |def lock_function_once(f, flag):                    |\n",
      "|                 |    def wrapper(*args, **kwargs):                   |\n",
      "|                 |        if not flag.invoked:                        |\n",
      "|                 |            with lock:                              |\n",
      "|                 |                if not flag.invoked:                |\n",
      "|                 |                    result = f(*args, **kwargs)     |\n",
      "|                 |                    flag.invoked = True             |\n",
      "|                 |                    return result                   |\n",
      "|                 |                return f(*args, **kwargs)           |\n",
      "|                 |        return f(*args, **kwargs)                   |\n",
      "|                 |    return wrapper                                  |\n",
      "|                 |                                                    |\n",
      "|                 |                                                    |\n",
      "|                 |invoked = InvokedFlag()                             |\n",
      "|                 |                                                    |\n",
      "|                 |def compute(df):                                    |\n",
      "|                 |    return lock_function_once(func, invoked)(*[...  |\n",
      "|                 |                                                    |\n",
      "|                 |import pandas                                       |\n",
      "|                 |                                                    |\n",
      "|                 |compute._sf_vectorized_input = pandas.DataFrame     |\n",
      "|                 |                                                    |\n",
      "|imports          |[@HOL_DB.PUBLIC.ML_MODELS/predict_risk_score.jo...  |\n",
      "|handler          |compute                                             |\n",
      "|runtime_version  |3.8                                                 |\n",
      "|packages         |['joblib==1.1.0','scikit-learn==1.1.1','cacheto...  |\n",
      "------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "session.sql(\"DESC FUNCTION PREDICT_RISK_SCORE()\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "21ec29e6-9564-46c3-8fef-9bf418c40a27",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------------------\n",
      "|\"TARGET\"  |\"PREDICTION\"         |\n",
      "----------------------------------\n",
      "|1         |0.0                  |\n",
      "|0         |1.0                  |\n",
      "|0         |0.0                  |\n",
      "|0         |0.0                  |\n",
      "|1         |0.625                |\n",
      "|0         |0.6666666666666666   |\n",
      "|0         |0.8148148148148148   |\n",
      "|1         |0.4666666666666667   |\n",
      "|0         |0.5                  |\n",
      "|0         |0.4375               |\n",
      "|1         |0.5454545454545454   |\n",
      "|1         |0.0                  |\n",
      "|1         |0.8                  |\n",
      "|1         |1.0                  |\n",
      "|1         |0.46153846153846156  |\n",
      "|1         |0.6595744680851063   |\n",
      "|1         |0.75                 |\n",
      "|0         |0.0                  |\n",
      "|0         |0.0                  |\n",
      "|0         |0.0                  |\n",
      "----------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "test_data_sf.with_column(\n",
    "    'PREDICTION', \n",
    "    predict_risk_score(F.object_construct('*')))\\\n",
    ".select(\"TARGET\",\"PREDICTION\").show(20)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}