Shows how to dynamically create a database schema using SQLAlchemy and a table schema stored in a Python dictionary. For completeness, includes creating the original database (in this case using PostgreSQL).

DynamicSchemaGeneration.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### PostgreSQL DB Creation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "\n",
    "from psycopg2 import connect\n",
    "from psycopg2.extensions import ISOLATION_LEVEL_AUTOCOMMIT"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "{'dbname': 'postgres',\n",
    " 'host': 'localhost',\n",
    " 'password': '********',\n",
    " 'port': '5432',\n",
    " 'user': '********'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('psql_local_creds.json', 'r') as file:\n",
    "    creds = json.loads(file.read())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "conn = connect(**creds)\n",
    "\n",
    "new_dbname = 'newdb'\n",
    "\n",
    "conn.set_isolation_level(ISOLATION_LEVEL_AUTOCOMMIT)\n",
    "with conn.cursor() as cursor:\n",
    "    cursor.execute(f'CREATE DATABASE {new_dbname}')\n",
    "    cursor.execute(f'GRANT ALL PRIVILEGES ON DATABASE {new_dbname} TO {creds[\"user\"]}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SQLAlchemy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sqlalchemy import create_engine, Table, Column, Integer, String, Float, MetaData\n",
    "\n",
    "# Connection string will look like this:\n",
    "# postgresql+psycopg2://username:password@localhost/newdb\n",
    "conn_string = f\"postgresql+psycopg2://{creds['user']}:{creds['password']}@{creds['host']}/{new_dbname}\"\n",
    "\n",
    "engine = create_engine(conn_string, echo=False)\n",
    "metadata = MetaData()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "schema = {\n",
    "    'dimensions': ['field_one', 'field_two'],\n",
    "    'integers': ['int_one', 'int_two'],\n",
    "    'floats': ['only_float'],\n",
    "}\n",
    "\n",
    "\n",
    "columns = [Column(column, String) for column in schema['dimensions']]\n",
    "columns += [Column(column, Integer) for column in schema['integers']]\n",
    "columns += [Column(column, Float) for column in schema['floats']]\n",
    "\n",
    "table = Table('my_table', metadata, *columns)\n",
    "\n",
    "metadata.drop_all(engine)\n",
    "metadata.create_all(engine)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Validate table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext sql\n",
    "%config SqlMagic.autopandas=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Connected: vgadmin@newdb'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%sql $conn_string"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " * postgresql+psycopg2://vgadmin:ttsms123@localhost/newdb\n",
      " * vgadmin@newdb\n",
      "1 rows affected.\n"
     ]
    },
    {
     "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: []\n",
       "Index: []"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%sql INSERT INTO my_table VALUES ('a', 'b', 1, 2, 3.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " * postgresql+psycopg2://vgadmin:ttsms123@localhost/newdb\n",
      " * vgadmin@newdb\n",
      "1 rows affected.\n"
     ]
    }
   ],
   "source": [
    "df = %sql SELECT * FROM my_table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>field_one</th>\n",
       "      <th>field_two</th>\n",
       "      <th>int_one</th>\n",
       "      <th>int_two</th>\n",
       "      <th>only_float</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>a</td>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3.3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  field_one field_two  int_one  int_two  only_float\n",
       "0         a         b        1        2         3.3"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}