notebooks/7-sp100_weights_optimization_via_drl.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "import gymnasium as gym\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import random\n",
    "import torch\n",
    "from gymnasium import spaces\n",
    "from tqdm import trange\n",
    "from torch import nn\n",
    "from torch_geometric.data import Data, Batch\n",
    "\n",
    "from notebooks.datasets.SP100Stocks import SP100Stocks\n",
    "from notebooks.models import A3TGCN\n",
    "from notebooks.ppo import PPO"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T14:22:36.304944700Z",
     "start_time": "2024-05-13T14:22:14.240892200Z"
    }
   },
   "id": "initial_id",
   "execution_count": 1
  },
  {
   "cell_type": "markdown",
   "source": [
    "# S&P100 weights optimization via Deep Reinforcement Learning\n",
    "The goal of this notebook is to build an agent capable of learning the optimal weights for the S&P100 index that give the best returns. The agent will try to outperform an equally weighted portfolio, meaning that each stock has the same weight.\n",
    "## Important note:\n",
    "### Unfortunately, this task uses high computational power and time, and I do not have the necessary resources to train the model. The code is provided as an example of how to implement a reinforcement learning agent to optimize the stock weights of a portfolio. The code has been tested for few epochs, which were not enough."
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "66fe37191902c89"
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Loading the data\n",
    "The data from the custom PyG dataset for containing the historical prices and the graph structure of the stocks is loaded.\n",
    "For this task, a longer time series than the previous ones is needed. Here 200 days of data is considered, which corresponds to 40 weeks."
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "db076977833be248"
  },
  {
   "cell_type": "code",
   "outputs": [
    {
     "data": {
      "text/plain": "(SP100Stocks(1208),\n Data(x=[100, 5, 25], edge_index=[2, 524], y=[100, 1], edge_weight=[524], close_price=[100, 25], close_price_y=[100, 1]))"
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "episode_duration = 200\n",
    "seq_len = 25\n",
    "dataset = SP100Stocks(past_window=seq_len)\n",
    "dataset, dataset[0]"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T14:22:38.075242300Z",
     "start_time": "2024-05-13T14:22:36.306940300Z"
    }
   },
   "id": "42734ad9d172b18",
   "execution_count": 2
  },
  {
   "cell_type": "markdown",
   "source": [
    "The data is split into a training and testing set. The testing contains the last `seq_len` days of the dataset. The training set contains the rest of the data."
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "f6e9b0b8a7014147"
  },
  {
   "cell_type": "code",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Creating training episodes: 100%|██████████| 101/101 [00:28<00:00,  3.58it/s]\n"
     ]
    },
    {
     "data": {
      "text/plain": "(101, SP100Stocks(200), SP100Stocks(200))"
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train_episodes has the shape (n_episodes, episode_duration, n_stocks, n_features, seq_len)\n",
    "train_episodes = [\n",
    "\tdataset[i:i+episode_duration] for i in trange(0, len(dataset)-episode_duration, 10, desc='Creating training episodes', position=0)\n",
    "]\n",
    "random.shuffle(train_episodes)\n",
    "\n",
    "test_episode = dataset[-episode_duration:]\n",
    "len(train_episodes), train_episodes[0], test_episode"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T14:23:06.857121300Z",
     "start_time": "2024-05-13T14:22:38.075242300Z"
    }
   },
   "id": "f85f8300c768b01",
   "execution_count": 3
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Reinforcement Learning Environment\n",
    "The environment is defined using the Gymnasium API.\n",
    "- State\n",
    "\n",
    "The observations are the graphs of the stocks at each time step, where nodes contain data on the last `n` days (variation, price, RSI, ...).\n",
    "- Action\n",
    "\n",
    "The action space is a continuous space of size (n_stocks,) where the agent can choose the weights for each stock. We need $\\sum_{i=1}^{n} a_i = 1$, where $a_i$ is the weight of stock $i$. A simple way to achieve this is to apply a softmax on the received action.\n",
    "- Reward\n",
    "\n",
    "The goal of the agent is to beat the baseline, which is an equally weighted portfolio. The reward is a function of the difference between the agent's performance and the baseline's performance. The reward is calculated with a sigmoid function as follows:\n",
    "$$\n",
    "r = \\left( \\frac{1}{1 + e^{b - a}} - 0.5 \\right) \\times c\n",
    "$$\n",
    "where $b$ is the baseline performance, $a$ is the agent's performance, and $c$ is a reward coefficient."
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "923bbdb1dccfa115"
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "class SP100Env(gym.Env):\n",
    "\tmetadata = {'render.modes': ['human']}\n",
    "\tdef __init__(self, n_stocks: int, n_edges: int, n_features, seq_len: int, train_episodes: list[SP100Stocks], test_episode: SP100Stocks, render_mode: str = None, reward_coef: float = 50):\n",
    "\t\tsuper(SP100Env, self).__init__()\n",
    "\t\tself.n_stocks: int = n_stocks\n",
    "\t\tself.n_features: int = n_features\n",
    "\t\tself.seq_len: int = seq_len\n",
    "\t\tself.episode_length: int = len(train_episodes[0]) - 1  # -1 because the next day is needed to calculate the reward\n",
    "\t\tself.reward_coef: float = reward_coef\n",
    "\t\t\n",
    "\t\tself.train_episodes: list[SP100Stocks] = train_episodes\n",
    "\t\tself.test_episode: list[SP100Stocks] = test_episode\n",
    "\t\t\n",
    "\t\tself.current_episode: SP100Stocks = train_episodes[0]\n",
    "\t\tself.current_step: int = 0\n",
    "\t\t\n",
    "\t\tassert render_mode is None or render_mode in self.metadata[\"render_modes\"]\n",
    "\t\tself.render_mode: str = render_mode\n",
    "\n",
    "\t\tself.action_space: spaces.Box = spaces.Box(low=0, high=1, shape=(n_stocks,), dtype=np.float32)\n",
    "\t\tself.observation_space: spaces.Dict = spaces.Dict({\n",
    "\t\t\t'x': spaces.Box(low=-np.inf, high=np.inf, shape=(n_stocks, n_features, seq_len), dtype=np.float32),\n",
    "\t\t\t'edge_index': spaces.Box(low=0, high=n_stocks-1, shape=(2, n_edges), dtype=np.int64),\n",
    "\t    'edge_weight': spaces.Box(low=-np.inf, high=np.inf, shape=(n_edges,), dtype=np.float32),\n",
    "\t\t})\n",
    "\t\t\n",
    "\t\tself.baseline_weights: np.ndarray = np.ones(n_stocks) / n_stocks\n",
    "\t\tself.current_baseline_performance: float = 1.\n",
    "\t\tself.current_agent_performance: float = 1.\n",
    "\t\t\n",
    "\tdef reset(self, seed=None, options=None) -> dict:\n",
    "\t\tif options and options.get(\"test\", False):\n",
    "\t\t\tself.current_episode = self.test_episode\n",
    "\t\telse:\n",
    "\t\t\tself.current_episode = random.choice(self.train_episodes)\n",
    "\t\tself.current_step = 0\n",
    "\t\tself.current_baseline_performance = 1.\n",
    "\t\tself.current_agent_performance = 1.\n",
    "\t\t\n",
    "\t\tself.render()\n",
    "\t\treturn self._get_observation(), {}\n",
    "\t\n",
    "\tdef step(self, action: np.ndarray) -> tuple[dict, float, bool, dict]:\n",
    "\t\taction = np.exp(action) / np.sum(np.exp(action))  # softmax to ensure weights sum to 1\n",
    "\t\tassert self.action_space.contains(action)\n",
    "\t\tassert action.shape == (self.n_stocks,)\n",
    "\t\t\n",
    "\t\t\n",
    "\t\tnext_day_variations = self.current_episode[self.current_step + 1].x[:, 2, -1].numpy() / 100  # variations are in percentage\n",
    "\t\t\n",
    "\t\tbaseline_variations = np.sum(next_day_variations * self.baseline_weights)\n",
    "\t\tagent_variations = np.sum(next_day_variations * action)\n",
    "\t\t\n",
    "\t\tself.current_baseline_performance *= (1 + baseline_variations)\n",
    "\t\tself.current_agent_performance *= (1 + agent_variations)\n",
    "\t\t\t\t\n",
    "\t\tself.current_step += 1\n",
    "\t\tobservation = self._get_observation()\n",
    "\t\treward = self._get_reward(baseline_variations, agent_variations)\n",
    "\t\tterminated = self.current_step >= self.episode_length\n",
    "\t\t\n",
    "\t\tself.render()\n",
    "\t\t\n",
    "\t\tif terminated:\n",
    "\t\t\tself.reset()\n",
    "\t\t\n",
    "\t\treturn observation, reward, terminated, False, {}\n",
    "\t\n",
    "\tdef render(self) -> None:\n",
    "\t\tif self.render_mode == 'human':\n",
    "\t\t\tprint(f'Current step: {self.current_step}/{self.episode_length}')\n",
    "\t\t\tprint(f'Baseline performance: {self.current_baseline_performance}')\n",
    "\t\t\tprint(f'Agent performance: {self.current_agent_performance}')\n",
    "\n",
    "\tdef close(self):\n",
    "\t\tpass\n",
    "\t\n",
    "\tdef _get_observation(self) -> dict:\n",
    "\t\treturn {\n",
    "\t\t\t'x': self.current_episode[self.current_step].x.numpy(),\n",
    "\t\t\t'edge_index': self.current_episode[self.current_step].edge_index.numpy(),\n",
    "\t\t\t'edge_weight': self.current_episode[self.current_step].edge_weight.numpy(),\n",
    "\t\t}\n",
    "\t\n",
    "\tdef _get_reward(self, step_baseline_performance: float, step_agent_performance: float) -> float:\n",
    "\t\treturn (1 / (1 + np.exp(step_baseline_performance - step_agent_performance)) - 1) * self.reward_coef"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T15:17:24.887561Z",
     "start_time": "2024-05-13T15:17:24.781485600Z"
    }
   },
   "id": "ab24cedc19d048f1",
   "execution_count": 20
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "n_stocks, n_edges, n_features = dataset[0].x.shape[0], dataset[0].edge_index.shape[1], dataset[0].x.shape[1]\n",
    "\n",
    "env = SP100Env(n_stocks=n_stocks, n_edges=n_edges,n_features=n_features, seq_len=seq_len, train_episodes=train_episodes, test_episode=test_episode)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T15:17:26.063996300Z",
     "start_time": "2024-05-13T15:17:24.890678700Z"
    }
   },
   "id": "205f263d794a6482",
   "execution_count": 21
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Defining the agent\n",
    "The agent uses the A3T-GCN model. The action space is a continuous space of size (n_stocks,). An Actor-Critic model is used. The actor will output a Gaussian distribution with a mean and a standard deviation of size (n_stocks,). The critic will output the value of the state."
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "a8fefc37273cc016"
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "class Actor(nn.Module):\n",
    "\tdef __init__(self, n_stocks: int, n_features: int, hidden_size: int, layers_nb: int):\n",
    "\t\tsuper(Actor, self).__init__()\n",
    "\t\tself.mean = A3TGCN(n_features, 1, hidden_size, layers_nb)\n",
    "\t\tself.logstd = nn.Parameter(torch.zeros(n_stocks))\n",
    "\t\n",
    "\tdef forward(self, obs: Data | Batch) -> tuple[torch.tensor, torch.tensor]:\n",
    "\t\treturn self.mean(obs.x, obs.edge_index, obs.edge_weight).view(obs.num_graphs, -1), torch.exp(self.logstd)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T15:17:26.069736Z",
     "start_time": "2024-05-13T15:17:26.055977300Z"
    }
   },
   "id": "a1cf2d9e5d53e219",
   "execution_count": 22
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "class Critic(nn.Module):\n",
    "\tdef __init__(self, n_stocks: int, n_features: int, hidden_size: int, layers_nb: int):\n",
    "\t\tsuper(Critic, self).__init__()\n",
    "\t\tself.value = A3TGCN(n_features, 1, hidden_size, layers_nb)\n",
    "\t\tself.out = nn.Linear(n_stocks, 1)\n",
    "\t\n",
    "\tdef forward(self, obs: Data | Batch) -> torch.tensor:\n",
    "\t\tvalues = self.value(obs.x, obs.edge_index, obs.edge_weight).view(obs.num_graphs, -1)\n",
    "\t\treturn self.out(values)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T15:17:26.070734300Z",
     "start_time": "2024-05-13T15:17:26.062775700Z"
    }
   },
   "id": "97ec4f01dfd98f76",
   "execution_count": 23
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Instantiating the agent & algorithm"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "c08b15dbbbe6f990"
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "hidden_size = 16\n",
    "actor_lr, critic_lr = 5e-3, 1e-2\n",
    "\n",
    "actor = Actor(n_stocks, n_features, hidden_size, 2)\n",
    "critic = Critic(n_stocks, n_features, hidden_size, 2)\n",
    "\n",
    "ppo = PPO(config={\n",
    "\t\"env_fn\": lambda: SP100Env(n_stocks=n_stocks, n_edges=n_edges, n_features=n_features, seq_len=seq_len, train_episodes=train_episodes, test_episode=test_episode),\n",
    "\t\"num_envs\": 2,\n",
    "\t\n",
    "\t\"actor_model\": actor,\n",
    "\t\"critic_model\": critic,\n",
    "\t\n",
    "\t\"horizon\": 16,\n",
    "})"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T15:17:26.112259900Z",
     "start_time": "2024-05-13T15:17:26.072850600Z"
    }
   },
   "id": "c7b6cae645f5db30",
   "execution_count": 24
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Training"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "29c8ca380b033e88"
  },
  {
   "cell_type": "code",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "==== STARTING TRAINING ====\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "PPO Training: 100%|██████████| 1000/1000 [19:14<00:00,  1.15s/it] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "==== TRAINING COMPLETE ====\n"
     ]
    }
   ],
   "source": [
    "ppo.train(max_steps=1_000, task_title=\"S&P100WeightsOptimization\")"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T15:38:25.247440100Z",
     "start_time": "2024-05-13T15:17:26.109052800Z"
    }
   },
   "id": "c0e03f2c91c5226",
   "execution_count": 25
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Evaluation\n",
    "The trained agent is tested on the testing set."
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "95ba72b7f327788b"
  },
  {
   "cell_type": "code",
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 1000x500 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "agent_performance, baseline_performance = [], []\n",
    "\n",
    "obs, _ = env.reset()\n",
    "done = False\n",
    "\n",
    "while not done:\n",
    "\tagent_performance.append(env.current_agent_performance)\n",
    "\tbaseline_performance.append(env.current_baseline_performance)\n",
    "\n",
    "\taction = ppo.compute_single_action(obs, {})\n",
    "\tobs, _, done, _, _ = env.step(action)\n",
    "\t\t\n",
    "plt.figure(figsize=(10, 5))\n",
    "plt.plot(agent_performance, label='Agent')\n",
    "plt.plot(baseline_performance, label='Baseline')\n",
    "plt.xlabel('Days')\n",
    "plt.ylabel('Portfolio performance')\n",
    "plt.title('Random agent vs Trained agent vs Baseline performance')\n",
    "plt.legend()\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-05-13T15:39:08.640944700Z",
     "start_time": "2024-05-13T15:38:25.253813Z"
    }
   },
   "id": "1fc92b3528ce18af",
   "execution_count": 26
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}