Os 400 workshop

workshops/open-source-l400/prep_sql_data.ipynb

@@ -0,0 +1,344 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Collecting Faker\n",
+      "  Downloading Faker-30.0.0-py3-none-any.whl.metadata (15 kB)\n",
+      "Requirement already satisfied: python-dateutil>=2.4 in /Users/jagdsoni/Desktop/TFC/code/bedrock-oss-workshop/.venv/lib/python3.11/site-packages (from Faker) (2.9.0.post0)\n",
+      "Requirement already satisfied: six>=1.5 in /Users/jagdsoni/Desktop/TFC/code/bedrock-oss-workshop/.venv/lib/python3.11/site-packages (from python-dateutil>=2.4->Faker) (1.16.0)\n",
+      "Downloading Faker-30.0.0-py3-none-any.whl (1.8 MB)\n",
+      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.8/1.8 MB\u001b[0m \u001b[31m19.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+      "\u001b[?25hInstalling collected packages: Faker\n",
+      "Successfully installed Faker-30.0.0\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install Faker"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Sample data has been generated and inserted into the SQLite database.\n"
+     ]
+    }
+   ],
+   "source": [
+    "import random\n",
+    "from datetime import datetime, timedelta\n",
+    "from faker import Faker\n",
+    "import sqlite3\n",
+    "import math\n",
+    "\n",
+    "# Initialize Faker\n",
+    "fake = Faker()\n",
+    "\n",
+    "# Get today's date\n",
+    "today = datetime.now().date()\n",
+    "\n",
+    "# City data with coordinates (latitude, longitude)\n",
+    "city_data = {\n",
+    "    \"New York\": (40.7128, -74.0060),\n",
+    "    \"Los Angeles\": (34.0522, -118.2437),\n",
+    "    \"Chicago\": (41.8781, -87.6298),\n",
+    "    \"Las Vegas\": (36.1699, -115.1398),\n",
+    "    \"San Francisco\": (37.7749, -122.4194),\n",
+    "    \"Orlando\": (28.5383, -81.3792),\n",
+    "    \"Miami\": (25.7617, -80.1918),\n",
+    "    \"Washington\": (38.9072, -77.0369),\n",
+    "    \"New Orleans\": (29.9511, -90.0715),\n",
+    "    \"Boston\": (42.3601, -71.0589),\n",
+    "    \"Seattle\": (47.6062, -122.3321),\n",
+    "    \"San Diego\": (32.7157, -117.1611),\n",
+    "    \"Honolulu\": (21.3069, -157.8583),\n",
+    "    \"Nashville\": (36.1627, -86.7816),\n",
+    "    \"Denver\": (39.7392, -104.9903),\n",
+    "    \"Austin\": (30.2672, -97.7431),\n",
+    "    \"Philadelphia\": (39.9526, -75.1652),\n",
+    "    \"Atlanta\": (33.7490, -84.3880),\n",
+    "    \"San Antonio\": (29.4241, -98.4936),\n",
+    "    \"Portland\": (45.5155, -122.6789),\n",
+    "    \"Paris\": (48.8566, 2.3522),\n",
+    "    \"London\": (51.5074, -0.1278),\n",
+    "    \"Rome\": (41.9028, 12.4964),\n",
+    "    \"Barcelona\": (41.3851, 2.1734),\n",
+    "    \"Amsterdam\": (52.3676, 4.9041),\n",
+    "    \"Berlin\": (52.5200, 13.4050),\n",
+    "    \"Vienna\": (48.2082, 16.3738),\n",
+    "    \"Prague\": (50.0755, 14.4378),\n",
+    "    \"Venice\": (45.4408, 12.3155),\n",
+    "    \"Florence\": (43.7696, 11.2558),\n",
+    "    \"Istanbul\": (41.0082, 28.9784),\n",
+    "    \"Madrid\": (40.4168, -3.7038),\n",
+    "    \"Lisbon\": (38.7223, -9.1393),\n",
+    "    \"Dublin\": (53.3498, -6.2603),\n",
+    "    \"Budapest\": (47.4979, 19.0402),\n",
+    "    \"Athens\": (37.9838, 23.7275),\n",
+    "    \"Edinburgh\": (55.9533, -3.1883),\n",
+    "    \"Copenhagen\": (55.6761, 12.5683),\n",
+    "    \"Stockholm\": (59.3293, 18.0686),\n",
+    "    \"Brussels\": (50.8503, 4.3517),\n",
+    "    \"Zurich\": (47.3769, 8.5417),\n",
+    "    \"Milan\": (45.4642, 9.1900),\n",
+    "    \"Munich\": (48.1351, 11.5820),\n",
+    "    \"Seville\": (37.3891, -5.9845),\n",
+    "}\n",
+    "\n",
+    "\n",
+    "def calculate_distance(lat1, lon1, lat2, lon2):\n",
+    "    R = 6371  # Earth's radius in kilometers\n",
+    "\n",
+    "    lat1, lon1, lat2, lon2 = map(math.radians, [lat1, lon1, lat2, lon2])\n",
+    "    dlat = lat2 - lat1\n",
+    "    dlon = lon2 - lon1\n",
+    "\n",
+    "    a = (\n",
+    "        math.sin(dlat / 2) ** 2\n",
+    "        + math.cos(lat1) * math.cos(lat2) * math.sin(dlon / 2) ** 2\n",
+    "    )\n",
+    "    c = 2 * math.asin(math.sqrt(a))\n",
+    "    distance = R * c\n",
+    "\n",
+    "    return distance\n",
+    "\n",
+    "\n",
+    "def calculate_flight_duration(distance):\n",
+    "    # Assume average speed of 800 km/h and add 30 minutes for takeoff and landing\n",
+    "    duration_hours = distance / 800\n",
+    "    duration_minutes = int(duration_hours * 60) + 30\n",
+    "    return timedelta(minutes=duration_minutes)\n",
+    "\n",
+    "\n",
+    "def calculate_flight_price(distance):\n",
+    "    # Base price of $50 plus $0.1 per km\n",
+    "    base_price = 50\n",
+    "    price_per_km = 0.1\n",
+    "    return round(base_price + (distance * price_per_km), 2)\n",
+    "\n",
+    "\n",
+    "def generate_user(user_id):\n",
+    "    name = fake.name()\n",
+    "    age = random.randint(18, 80)\n",
+    "    home_location = random.choice(list(city_data.keys()))\n",
+    "\n",
+    "    return (user_id, name, age, home_location)\n",
+    "\n",
+    "\n",
+    "def generate_flight_booking(booking_id, user_id, user_name):\n",
+    "    origin = random.choice(list(city_data.keys()))\n",
+    "    destination = random.choice([city for city in city_data.keys() if city != origin])\n",
+    "\n",
+    "    origin_coords = city_data[origin]\n",
+    "    dest_coords = city_data[destination]\n",
+    "\n",
+    "    distance = calculate_distance(\n",
+    "        origin_coords[0], origin_coords[1], dest_coords[0], dest_coords[1]\n",
+    "    )\n",
+    "    flight_duration = calculate_flight_duration(distance)\n",
+    "    price = calculate_flight_price(distance)\n",
+    "\n",
+    "    # Generate a random booking date within the next 15 days\n",
+    "    booking_date = today + timedelta(days=random.randint(0, 14))\n",
+    "\n",
+    "    # Generate departure date between 1 and 30 days after the booking date\n",
+    "    departure_date = booking_date + timedelta(days=random.randint(1, 30))\n",
+    "\n",
+    "    # Generate departure time\n",
+    "    departure_time = f\"{random.randint(0, 23):02d}:{random.randint(0, 59):02d}\"\n",
+    "\n",
+    "    # Calculate arrival date and time\n",
+    "    departure_datetime = datetime.combine(\n",
+    "        departure_date, datetime.strptime(departure_time, \"%H:%M\").time()\n",
+    "    )\n",
+    "    arrival_datetime = departure_datetime + flight_duration\n",
+    "\n",
+    "    return (\n",
+    "        booking_id,\n",
+    "        user_id,\n",
+    "        user_name,\n",
+    "        origin,\n",
+    "        destination,\n",
+    "        price,\n",
+    "        int(flight_duration.total_seconds() // 60),  # Convert to minutes\n",
+    "        departure_date.strftime(\"%Y-%m-%d\"),\n",
+    "        departure_time,\n",
+    "        arrival_datetime.strftime(\"%Y-%m-%d\"),\n",
+    "        arrival_datetime.strftime(\"%H:%M\"),\n",
+    "        round(distance, 2),\n",
+    "        booking_date.strftime(\"%Y-%m-%d\"),\n",
+    "    )\n",
+    "\n",
+    "\n",
+    "def generate_hotel_booking(booking_id, user_id, user_name):\n",
+    "    city = random.choice(list(city_data.keys()))\n",
+    "    hotel_name = fake.company() + \" Hotel\"\n",
+    "    check_in_date = fake.date_between(start_date=\"+1d\", end_date=\"+60d\")\n",
+    "    nights = random.randint(1, 14)\n",
+    "    check_out_date = check_in_date + timedelta(days=nights)\n",
+    "    price_per_night = round(random.uniform(50, 500), 2)\n",
+    "    total_price = round(price_per_night * nights, 2)\n",
+    "    num_guests = random.randint(1, 4)\n",
+    "    room_type = random.choice([\"Single\", \"Double\", \"Suite\", \"Deluxe\"])\n",
+    "\n",
+    "    return (\n",
+    "        booking_id,\n",
+    "        user_id,\n",
+    "        user_name,\n",
+    "        city,\n",
+    "        hotel_name,\n",
+    "        check_in_date.strftime(\"%Y-%m-%d\"),\n",
+    "        check_out_date.strftime(\"%Y-%m-%d\"),\n",
+    "        nights,\n",
+    "        price_per_night,\n",
+    "        total_price,\n",
+    "        num_guests,\n",
+    "        room_type,\n",
+    "    )\n",
+    "\n",
+    "\n",
+    "def create_database():\n",
+    "    conn = sqlite3.connect(\"travel_bookings.db\")\n",
+    "    cursor = conn.cursor()\n",
+    "\n",
+    "    cursor.execute(\n",
+    "        \"\"\"\n",
+    "    CREATE TABLE IF NOT EXISTS users (\n",
+    "        user_id INTEGER PRIMARY KEY,\n",
+    "        name TEXT,\n",
+    "        age INTEGER,\n",
+    "        home_location TEXT\n",
+    "    )\n",
+    "    \"\"\"\n",
+    "    )\n",
+    "\n",
+    "    cursor.execute(\n",
+    "        \"\"\"\n",
+    "    CREATE TABLE IF NOT EXISTS flight_bookings (\n",
+    "        booking_id INTEGER PRIMARY KEY,\n",
+    "        user_id INTEGER,\n",
+    "        user_name TEXT,\n",
+    "        origin TEXT,\n",
+    "        destination TEXT,\n",
+    "        price REAL,\n",
+    "        flight_duration INTEGER,\n",
+    "        departure_date TEXT,\n",
+    "        departure_time TEXT,\n",
+    "        arrival_date TEXT,\n",
+    "        arrival_time TEXT,\n",
+    "        distance REAL,\n",
+    "        booking_date TEXT,\n",
+    "        FOREIGN KEY (user_id) REFERENCES users (user_id)\n",
+    "    )\n",
+    "    \"\"\"\n",
+    "    )\n",
+    "\n",
+    "    cursor.execute(\n",
+    "        \"\"\"\n",
+    "    CREATE TABLE IF NOT EXISTS hotel_bookings (\n",
+    "        booking_id INTEGER PRIMARY KEY,\n",
+    "        user_id INTEGER,\n",
+    "        user_name TEXT,\n",
+    "        city TEXT,\n",
+    "        hotel_name TEXT,\n",
+    "        check_in_date TEXT,\n",
+    "        check_out_date TEXT,\n",
+    "        nights INTEGER,\n",
+    "        price_per_night REAL,\n",
+    "        total_price REAL,\n",
+    "        num_guests INTEGER,\n",
+    "        room_type TEXT,\n",
+    "        FOREIGN KEY (user_id) REFERENCES users (user_id)\n",
+    "    )\n",
+    "    \"\"\"\n",
+    "    )\n",
+    "\n",
+    "    conn.commit()\n",
+    "    return conn, cursor\n",
+    "\n",
+    "\n",
+    "def insert_sample_data(\n",
+    "    conn, cursor, num_users, num_flight_bookings, num_hotel_bookings\n",
+    "):\n",
+    "    # Generate and insert user data\n",
+    "    users = [generate_user(i + 1) for i in range(num_users)]\n",
+    "    cursor.executemany(\n",
+    "        \"\"\"\n",
+    "    INSERT INTO users (user_id, name, age, home_location)\n",
+    "    VALUES (?, ?, ?, ?)\n",
+    "    \"\"\",\n",
+    "        users,\n",
+    "    )\n",
+    "\n",
+    "    # Generate and insert flight bookings\n",
+    "    for i in range(num_flight_bookings):\n",
+    "        user = random.choice(users)\n",
+    "        flight_booking = generate_flight_booking(i + 1, user[0], user[1])\n",
+    "        cursor.execute(\n",
+    "            \"\"\"\n",
+    "        INSERT INTO flight_bookings (booking_id, user_id, user_name, origin, destination, price, flight_duration, departure_date, departure_time, arrival_date, arrival_time, distance, booking_date)\n",
+    "        VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)\n",
+    "        \"\"\",\n",
+    "            flight_booking,\n",
+    "        )\n",
+    "\n",
+    "    # Generate and insert hotel bookings\n",
+    "    for i in range(num_hotel_bookings):\n",
+    "        user = random.choice(users)\n",
+    "        hotel_booking = generate_hotel_booking(i + 1, user[0], user[1])\n",
+    "        cursor.execute(\n",
+    "            \"\"\"\n",
+    "        INSERT INTO hotel_bookings (booking_id, user_id, user_name, city, hotel_name, check_in_date, check_out_date, nights, price_per_night, total_price, num_guests, room_type)\n",
+    "        VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)\n",
+    "        \"\"\",\n",
+    "            hotel_booking,\n",
+    "        )\n",
+    "\n",
+    "    conn.commit()\n",
+    "\n",
+    "\n",
+    "if __name__ == \"__main__\":\n",
+    "    conn, cursor = create_database()\n",
+    "    insert_sample_data(\n",
+    "        conn, cursor, num_users=500, num_flight_bookings=1000, num_hotel_bookings=1000\n",
+    "    )\n",
+    "    conn.close()\n",
+    "    print(\"Sample data has been generated and inserted into the SQLite database.\")"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "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": "3.11.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}