{ "cells": [ { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "#| default_exp models.tcn" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "#| hide\n", "%load_ext autoreload\n", "%autoreload 2" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# TCN" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "For long time in deep learning, sequence modelling was synonymous with recurrent networks, yet several papers have shown that simple convolutional architectures can outperform canonical recurrent networks like LSTMs by demonstrating longer effective memory. By skipping temporal connections the causal convolution filters can be applied to larger time spans while remaining computationally efficient.\n", "\n", "The predictions are obtained by transforming the hidden states into contexts $\\mathbf{c}_{[t+1:t+H]}$, that are decoded and adapted into $\\mathbf{\\hat{y}}_{[t+1:t+H],[q]}$ through MLPs.\n", "\n", "\\begin{align}\n", " \\mathbf{h}_{t} &= \\textrm{TCN}([\\mathbf{y}_{t},\\mathbf{x}^{(h)}_{t},\\mathbf{x}^{(s)}], \\mathbf{h}_{t-1})\\\\\n", "\\mathbf{c}_{[t+1:t+H]}&=\\textrm{Linear}([\\mathbf{h}_{t}, \\mathbf{x}^{(f)}_{[:t+H]}]) \\\\ \n", "\\hat{y}_{\\tau,[q]}&=\\textrm{MLP}([\\mathbf{c}_{\\tau},\\mathbf{x}^{(f)}_{\\tau}])\n", "\\end{align}\n", "\n", "where $\\mathbf{h}_{t}$, is the hidden state for time $t$, $\\mathbf{y}_{t}$ is the input at time $t$ and $\\mathbf{h}_{t-1}$ is the hidden state of the previous layer at $t-1$, $\\mathbf{x}^{(s)}$ are static exogenous inputs, $\\mathbf{x}^{(h)}_{t}$ historic exogenous, $\\mathbf{x}^{(f)}_{[:t+H]}$ are future exogenous available at the time of the prediction.\n", "\n", "**References**
\n", "-[van den Oord, A., Dieleman, S., Zen, H., Simonyan, K., Vinyals, O., Graves, A., Kalchbrenner, N., Senior, A. W., & Kavukcuoglu, K. (2016). Wavenet: A generative model for raw audio. Computing Research Repository, abs/1609.03499. URL: http://arxiv.org/abs/1609.03499. arXiv:1609.03499.](https://arxiv.org/abs/1609.03499)
\n", "-[Shaojie Bai, Zico Kolter, Vladlen Koltun. (2018). An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling. Computing Research Repository, abs/1803.01271. URL: https://arxiv.org/abs/1803.01271.](https://arxiv.org/abs/1803.01271)
" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "![Figure 1. Visualization of a stack of dilated causal convolutional layers.](imgs_models/tcn.png)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "#| export\n", "from typing import List, Optional\n", "\n", "import torch\n", "import torch.nn as nn\n", "\n", "from neuralforecast.losses.pytorch import MAE\n", "from neuralforecast.common._base_recurrent import BaseRecurrent\n", "from neuralforecast.common._modules import MLP, TemporalConvolutionEncoder" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "#| hide\n", "from nbdev.showdoc import show_doc\n", "\n", "import logging\n", "import warnings" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "#| export\n", "class TCN(BaseRecurrent):\n", " \"\"\" TCN\n", "\n", " Temporal Convolution Network (TCN), with MLP decoder.\n", " The historical encoder uses dilated skip connections to obtain efficient long memory,\n", " while the rest of the architecture allows for future exogenous alignment.\n", "\n", " **Parameters:**
\n", " `h`: int, forecast horizon.
\n", " `input_size`: int, maximum sequence length for truncated train backpropagation. Default -1 uses all history.
\n", " `inference_input_size`: int, maximum sequence length for truncated inference. Default -1 uses all history.
\n", " `kernel_size`: int, size of the convolving kernel.
\n", " `dilations`: int list, ontrols the temporal spacing between the kernel points; also known as the à trous algorithm.
\n", " `encoder_hidden_size`: int=200, units for the TCN's hidden state size.
\n", " `encoder_activation`: str=`tanh`, type of TCN activation from `tanh` or `relu`.
\n", " `context_size`: int=10, size of context vector for each timestamp on the forecasting window.
\n", " `decoder_hidden_size`: int=200, size of hidden layer for the MLP decoder.
\n", " `decoder_layers`: int=2, number of layers for the MLP decoder.
\n", " `futr_exog_list`: str list, future exogenous columns.
\n", " `hist_exog_list`: str list, historic exogenous columns.
\n", " `stat_exog_list`: str list, static exogenous columns.
\n", " `loss`: PyTorch module, instantiated train loss class from [losses collection](https://nixtla.github.io/neuralforecast/losses.pytorch.html).
\n", " `max_steps`: int=1000, maximum number of training steps.
\n", " `learning_rate`: float=1e-3, Learning rate between (0, 1).
\n", " `valid_batch_size`: int=None, number of different series in each validation and test batch.
\n", " `num_lr_decays`: int=-1, Number of learning rate decays, evenly distributed across max_steps.
\n", " `early_stop_patience_steps`: int=-1, Number of validation iterations before early stopping.
\n", " `val_check_steps`: int=100, Number of training steps between every validation loss check.
`batch_size`: int=32, number of differentseries in each batch.
\n", " `scaler_type`: str='robust', type of scaler for temporal inputs normalization see [temporal scalers](https://nixtla.github.io/neuralforecast/common.scalers.html).
\n", " `random_seed`: int=1, random_seed for pytorch initializer and numpy generators.
\n", " `num_workers_loader`: int=os.cpu_count(), workers to be used by `TimeSeriesDataLoader`.
\n", " `drop_last_loader`: bool=False, if True `TimeSeriesDataLoader` drops last non-full batch.
\n", " `alias`: str, optional, Custom name of the model.
\n", " `optimizer`: Subclass of 'torch.optim.Optimizer', optional, user specified optimizer instead of the default choice (Adam).
\n", " `optimizer_kwargs`: dict, optional, list of parameters used by the user specified `optimizer`.
\n", " `**trainer_kwargs`: int, keyword trainer arguments inherited from [PyTorch Lighning's trainer](https://pytorch-lightning.readthedocs.io/en/stable/api/pytorch_lightning.trainer.trainer.Trainer.html?highlight=trainer).
\n", " \"\"\"\n", " # Class attributes\n", " SAMPLING_TYPE = 'recurrent'\n", " \n", " def __init__(self,\n", " h: int,\n", " input_size: int = -1,\n", " inference_input_size: int = -1,\n", " kernel_size: int = 2,\n", " dilations: List[int] = [1, 2, 4, 8, 16],\n", " encoder_hidden_size: int = 200,\n", " encoder_activation: str = 'ReLU',\n", " context_size: int = 10,\n", " decoder_hidden_size: int = 200,\n", " decoder_layers: int = 2,\n", " futr_exog_list = None,\n", " hist_exog_list = None,\n", " stat_exog_list = None,\n", " loss=MAE(),\n", " valid_loss=None,\n", " max_steps: int = 1000,\n", " learning_rate: float = 1e-3,\n", " num_lr_decays: int = -1,\n", " early_stop_patience_steps: int =-1,\n", " val_check_steps: int = 100,\n", " batch_size: int = 32,\n", " valid_batch_size: Optional[int] = None,\n", " scaler_type: str ='robust',\n", " random_seed: int = 1,\n", " num_workers_loader = 0,\n", " drop_last_loader = False,\n", " optimizer = None,\n", " optimizer_kwargs = None,\n", " **trainer_kwargs):\n", " super(TCN, self).__init__(\n", " h=h,\n", " input_size=input_size,\n", " inference_input_size=inference_input_size,\n", " loss=loss,\n", " valid_loss=valid_loss,\n", " max_steps=max_steps,\n", " learning_rate=learning_rate,\n", " num_lr_decays=num_lr_decays,\n", " early_stop_patience_steps=early_stop_patience_steps,\n", " val_check_steps=val_check_steps,\n", " batch_size=batch_size,\n", " valid_batch_size=valid_batch_size,\n", " scaler_type=scaler_type,\n", " futr_exog_list=futr_exog_list,\n", " hist_exog_list=hist_exog_list,\n", " stat_exog_list=stat_exog_list,\n", " num_workers_loader=num_workers_loader,\n", " drop_last_loader=drop_last_loader,\n", " random_seed=random_seed,\n", " optimizer=optimizer,\n", " optimizer_kwargs=optimizer_kwargs,\n", " **trainer_kwargs\n", " )\n", "\n", " #----------------------------------- Parse dimensions -----------------------------------#\n", " # TCN\n", " self.kernel_size = kernel_size\n", " self.dilations = dilations\n", " self.encoder_hidden_size = encoder_hidden_size\n", " self.encoder_activation = encoder_activation\n", " \n", " # Context adapter\n", " self.context_size = context_size\n", "\n", " # MLP decoder\n", " self.decoder_hidden_size = decoder_hidden_size\n", " self.decoder_layers = decoder_layers\n", "\n", " self.futr_exog_size = len(self.futr_exog_list)\n", " self.hist_exog_size = len(self.hist_exog_list)\n", " self.stat_exog_size = len(self.stat_exog_list)\n", " \n", " # TCN input size (1 for target variable y)\n", " input_encoder = 1 + self.hist_exog_size + self.stat_exog_size\n", "\n", " \n", " #---------------------------------- Instantiate Model -----------------------------------#\n", " # Instantiate historic encoder\n", " self.hist_encoder = TemporalConvolutionEncoder(\n", " in_channels=input_encoder,\n", " out_channels=self.encoder_hidden_size,\n", " kernel_size=self.kernel_size, # Almost like lags\n", " dilations=self.dilations,\n", " activation=self.encoder_activation)\n", "\n", " # Context adapter\n", " self.context_adapter = nn.Linear(in_features=self.encoder_hidden_size + self.futr_exog_size * h,\n", " out_features=self.context_size * h)\n", "\n", " # Decoder MLP\n", " self.mlp_decoder = MLP(in_features=self.context_size + self.futr_exog_size,\n", " out_features=self.loss.outputsize_multiplier,\n", " hidden_size=self.decoder_hidden_size,\n", " num_layers=self.decoder_layers,\n", " activation='ReLU',\n", " dropout=0.0)\n", "\n", " def forward(self, windows_batch):\n", " \n", " # Parse windows_batch\n", " encoder_input = windows_batch['insample_y'] # [B, seq_len, 1]\n", " futr_exog = windows_batch['futr_exog']\n", " hist_exog = windows_batch['hist_exog']\n", " stat_exog = windows_batch['stat_exog']\n", "\n", " # Concatenate y, historic and static inputs\n", " # [B, C, seq_len, 1] -> [B, seq_len, C]\n", " # Contatenate [ Y_t, | X_{t-L},..., X_{t} | S ]\n", " batch_size, seq_len = encoder_input.shape[:2]\n", " if self.hist_exog_size > 0:\n", " hist_exog = hist_exog.permute(0,2,1,3).squeeze(-1) # [B, X, seq_len, 1] -> [B, seq_len, X]\n", " encoder_input = torch.cat((encoder_input, hist_exog), dim=2)\n", "\n", " if self.stat_exog_size > 0:\n", " stat_exog = stat_exog.unsqueeze(1).repeat(1, seq_len, 1) # [B, S] -> [B, seq_len, S]\n", " encoder_input = torch.cat((encoder_input, stat_exog), dim=2)\n", "\n", " # TCN forward\n", " hidden_state = self.hist_encoder(encoder_input) # [B, seq_len, tcn_hidden_state]\n", "\n", " if self.futr_exog_size > 0:\n", " futr_exog = futr_exog.permute(0,2,3,1)[:,:,1:,:] # [B, F, seq_len, 1+H] -> [B, seq_len, H, F]\n", " hidden_state = torch.cat(( hidden_state, futr_exog.reshape(batch_size, seq_len, -1)), dim=2)\n", "\n", " # Context adapter\n", " context = self.context_adapter(hidden_state)\n", " context = context.reshape(batch_size, seq_len, self.h, self.context_size)\n", "\n", " # Residual connection with futr_exog\n", " if self.futr_exog_size > 0:\n", " context = torch.cat((context, futr_exog), dim=-1)\n", "\n", " # Final forecast\n", " output = self.mlp_decoder(context)\n", " output = self.loss.domain_map(output)\n", " \n", " return output" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "show_doc(TCN)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "show_doc(TCN.fit, name='TCN.fit')" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "show_doc(TCN.predict, name='TCN.predict')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Usage Example" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "#| hide\n", "logging.getLogger(\"pytorch_lightning\").setLevel(logging.ERROR)\n", "warnings.filterwarnings(\"ignore\")" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "#| eval: false\n", "import numpy as np\n", "import pandas as pd\n", "import pytorch_lightning as pl\n", "import matplotlib.pyplot as plt\n", "\n", "from neuralforecast import NeuralForecast\n", "from neuralforecast.models import TCN\n", "from neuralforecast.losses.pytorch import GMM, MQLoss, DistributionLoss\n", "from neuralforecast.utils import AirPassengersPanel, AirPassengersStatic\n", "from neuralforecast.tsdataset import TimeSeriesDataset, TimeSeriesLoader\n", "\n", "Y_train_df = AirPassengersPanel[AirPassengersPanel.ds=AirPassengersPanel['ds'].values[-12]].reset_index(drop=True) # 12 test\n", "\n", "fcst = NeuralForecast(\n", " models=[TCN(h=12,\n", " input_size=-1,\n", " #loss=DistributionLoss(distribution='Normal', level=[80, 90]),\n", " loss=GMM(n_components=7, return_params=True, level=[80,90]),\n", " learning_rate=5e-4,\n", " kernel_size=2,\n", " dilations=[1,2,4,8,16],\n", " encoder_hidden_size=128,\n", " context_size=10,\n", " decoder_hidden_size=128,\n", " decoder_layers=2,\n", " max_steps=500,\n", " scaler_type='robust',\n", " futr_exog_list=['y_[lag12]'],\n", " hist_exog_list=None,\n", " stat_exog_list=['airline1'],\n", " )\n", " ],\n", " freq='M'\n", ")\n", "fcst.fit(df=Y_train_df, static_df=AirPassengersStatic)\n", "forecasts = fcst.predict(futr_df=Y_test_df)\n", "\n", "# Plot quantile predictions\n", "Y_hat_df = forecasts.reset_index(drop=False).drop(columns=['unique_id','ds'])\n", "plot_df = pd.concat([Y_test_df, Y_hat_df], axis=1)\n", "plot_df = pd.concat([Y_train_df, plot_df])\n", "\n", "plot_df = plot_df[plot_df.unique_id=='Airline1'].drop('unique_id', axis=1)\n", "plt.plot(plot_df['ds'], plot_df['y'], c='black', label='True')\n", "plt.plot(plot_df['ds'], plot_df['TCN-median'], c='blue', label='median')\n", "plt.fill_between(x=plot_df['ds'][-12:], \n", " y1=plot_df['TCN-lo-90'][-12:].values,\n", " y2=plot_df['TCN-hi-90'][-12:].values,\n", " alpha=0.4, label='level 90')\n", "plt.legend()\n", "plt.grid()\n", "plt.plot()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "python3", "language": "python", "name": "python3" } }, "nbformat": 4, "nbformat_minor": 4 }