{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Recurrent Neural Networks Models\n", "In this notebook, we show an example of how RNNs can be used with darts.\n", "If you are new to darts, we recommend you first follow the [quick start](https://unit8co.github.io/darts/quickstart/00-quickstart.html) notebook." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "# fix python path if working locally\n", "from utils import fix_pythonpath_if_working_locally\n", "\n", "fix_pythonpath_if_working_locally()" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "%load_ext autoreload\n", "%autoreload 2\n", "%matplotlib inline" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "tags": [] }, "outputs": [], "source": [ "import warnings\n", "\n", "import matplotlib.pyplot as plt\n", "import numpy as np\n", "import pandas as pd\n", "\n", "from darts.dataprocessing.transformers import Scaler\n", "from darts.datasets import AirPassengersDataset, SunspotsDataset\n", "from darts.metrics import mape\n", "from darts.models import BlockRNNModel, ExponentialSmoothing, RNNModel\n", "from darts.utils.callbacks import TFMProgressBar\n", "from darts.utils.statistics import check_seasonality, plot_acf\n", "from darts.utils.timeseries_generation import datetime_attribute_timeseries\n", "\n", "warnings.filterwarnings(\"ignore\")\n", "import logging\n", "\n", "logging.disable(logging.CRITICAL)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Recurrent Models\n", "\n", "Darts includes two recurrent forecasting model classes: `RNNModel` and `BlockRNNModel`. \n", "\n", "`RNNModel` is fully recurrent in the sense that, at prediction time, an output is computed using these inputs:\n", "\n", "- the previous target value, which will be set to the last known target value for the first prediction,\n", " and for all other predictions it will be set to the previous prediction\n", "- the previous hidden state\n", "- the current covariates (if the model was trained with covariates)\n", "\n", "A prediction with forecasting horizon `n` thus is created in `n` iterations of `RNNModel` predictions and requires `n` future covariates to be known. This model is suited for forecasting problems where the target series is highly dependent on covariates that are known in advance.\n", "\n", "`BlockRNNModel` has a recurrent encoder stage, which encodes its input, and a fully-connected neural network decoder stage, which produces a prediction of length `output_chunk_length` based on the last hidden state of the encoder stage. Consequently, this model produces 'blocks' of forecasts and is restricted to looking at covariates with the same time index as the input target series." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Air Passenger Example\n", "This is a data set that is highly dependent on covariates. Knowing the month tells us a lot about the seasonal component, whereas the year determines the effect of the trend component. Both of these covariates are known in the future, and thus the `RNNModel` class is the preferred choice for this problem." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "# Read data:\n", "series = AirPassengersDataset().load().astype(np.float32)\n", "\n", "# Create training and validation sets:\n", "train, val = series.split_after(pd.Timestamp(\"19590101\"))\n", "\n", "# Normalize the time series (note: we avoid fitting the transformer on the validation set)\n", "transformer = Scaler()\n", "train_transformed = transformer.fit_transform(train)\n", "val_transformed = transformer.transform(val)\n", "series_transformed = transformer.transform(series)\n", "\n", "# create month and year covariate series\n", "year_series = datetime_attribute_timeseries(\n", " pd.date_range(start=series.start_time(), freq=series.freq_str, periods=1000),\n", " attribute=\"year\",\n", " one_hot=False,\n", " dtype=series.dtype,\n", ")\n", "year_series = Scaler().fit_transform(year_series)\n", "month_series = datetime_attribute_timeseries(\n", " year_series,\n", " attribute=\"month\",\n", " one_hot=True,\n", " dtype=series.dtype,\n", ")\n", "covariates = year_series.stack(month_series)\n", "cov_train, cov_val = covariates.split_after(pd.Timestamp(\"19590101\"))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's train an LSTM neural net. For using vanilla RNN or GRU instead, replace `'LSTM'` by `'RNN'` or `'GRU'`, respectively." ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "scrolled": true }, "outputs": [], "source": [ "my_model = RNNModel(\n", " input_chunk_length=14,\n", " training_length=20,\n", " model=\"LSTM\",\n", " hidden_dim=20,\n", " dropout=0,\n", " batch_size=16,\n", " n_epochs=300,\n", " optimizer_kwargs={\"lr\": 1e-3},\n", " model_name=\"Air_RNN\",\n", " log_tensorboard=True,\n", " random_state=42,\n", " force_reset=True,\n", " save_checkpoints=True,\n", " pl_trainer_kwargs={\"callbacks\": [TFMProgressBar(enable_train_bar_only=True)]},\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In what follows, we can just provide the whole `covariates` series as `future_covariates` argument to the model; the model will slice these covariates and use only what it needs in order to train on forecasting the target `train_transformed`:" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "tags": [] }, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "7e18d8e02b274b88955b6710a7ab90c2", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Training: 0it [00:00, ?it/s]" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [ "RNNModel(model=LSTM, hidden_dim=20, n_rnn_layers=1, dropout=0, training_length=20, input_chunk_length=14, batch_size=16, n_epochs=300, optimizer_kwargs={'lr': 0.001}, model_name=Air_RNN, log_tensorboard=True, random_state=42, force_reset=True, save_checkpoints=True, pl_trainer_kwargs={'callbacks': []})" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_model.fit(\n", " train_transformed,\n", " future_covariates=covariates,\n", " val_series=val_transformed,\n", " val_future_covariates=covariates,\n", " verbose=True,\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Look at predictions on the validation set\n", "Use the \"current\" model - i.e., the model at the end of the training procedure:" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "def eval_model(model):\n", " pred_series = model.predict(n=26, future_covariates=covariates)\n", " plt.figure(figsize=(8, 5))\n", " series_transformed.plot(label=\"actual\")\n", " pred_series.plot(label=\"forecast\")\n", " plt.title(f\"MAPE: {mape(pred_series, val_transformed):.2f}%\")\n", " plt.legend()\n", "\n", "\n", "eval_model(my_model)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Use the best model obtained over training, according to validation loss:" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "best_model = RNNModel.load_from_checkpoint(model_name=\"Air_RNN\", best=True)\n", "eval_model(best_model)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Backtesting\n", "Let's backtest our `RNN` model, to see how it performs at a forecast horizon of 6 months:" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "tags": [] }, "outputs": [], "source": [ "backtest_series = my_model.historical_forecasts(\n", " series_transformed,\n", " future_covariates=covariates,\n", " start=pd.Timestamp(\"19590101\"),\n", " forecast_horizon=6,\n", " retrain=False,\n", " verbose=True,\n", ")" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "MAPE: 2.71%\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plt.figure(figsize=(8, 5))\n", "series_transformed.plot(label=\"actual\")\n", "backtest_series.plot(label=\"backtest\")\n", "plt.legend()\n", "plt.title(\"Backtest, starting Jan 1959, 6-months horizon\")\n", "print(\n", " \"MAPE: {:.2f}%\".format(\n", " mape(\n", " transformer.inverse_transform(series_transformed),\n", " transformer.inverse_transform(backtest_series),\n", " )\n", " )\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Monthly sunspots\n", "Let's now try a more challenging time series; that of the monthly number of sunspots since 1749. First, we build the time series from the data, and check its periodicity." ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(True, np.int64(125))" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "series_sunspot = SunspotsDataset().load().astype(np.float32)\n", "\n", "series_sunspot.plot()\n", "check_seasonality(series_sunspot, max_lag=240)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_acf(series_sunspot, 125, max_lag=240) # ~11 years seasonality" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [], "source": [ "train_sp, val_sp = series_sunspot.split_after(pd.Timestamp(\"19401001\"))\n", "\n", "transformer_sunspot = Scaler()\n", "train_sp_transformed = transformer_sunspot.fit_transform(train_sp)\n", "val_sp_transformed = transformer_sunspot.transform(val_sp)\n", "series_sp_transformed = transformer_sunspot.transform(series_sunspot)" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "scrolled": true, "tags": [] }, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "5d68599a153c4affb8131b573ae709d0", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Training: 0it [00:00, ?it/s]" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [ "BlockRNNModel(output_chunk_shift=0, model=GRU, hidden_dim=10, n_rnn_layers=1, hidden_fc_sizes=None, dropout=0.1, activation=ReLU, use_static_covariates=True, input_chunk_length=125, output_chunk_length=36, batch_size=32, n_epochs=100, model_name=sun_GRU, nr_epochs_val_period=1, optimizer_kwargs={'lr': 0.001}, log_tensorboard=True, random_state=42, force_reset=True, pl_trainer_kwargs={'callbacks': []})" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "my_model_sun = BlockRNNModel(\n", " model=\"GRU\",\n", " input_chunk_length=125,\n", " output_chunk_length=36,\n", " hidden_dim=10,\n", " n_rnn_layers=1,\n", " batch_size=32,\n", " n_epochs=100,\n", " dropout=0.1,\n", " model_name=\"sun_GRU\",\n", " nr_epochs_val_period=1,\n", " optimizer_kwargs={\"lr\": 1e-3},\n", " log_tensorboard=True,\n", " random_state=42,\n", " force_reset=True,\n", " pl_trainer_kwargs={\"callbacks\": [TFMProgressBar(enable_train_bar_only=True)]},\n", ")\n", "\n", "my_model_sun.fit(train_sp_transformed, val_series=val_sp_transformed, verbose=True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To evaluate our model, we will simulate historic forecasting with a forecasting horizon of 3 years across the validation set. To speed things up, we will only look at every 10th forecast. For the sake of comparison, let's also fit an exponential smoothing model." ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "5b2e222d6ddb4b63aeaa1e520b04dae9", "version_major": 2, "version_minor": 0 }, "text/plain": [ "historical forecasts: 0%| | 0/49 [00:00" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "val_sp_transformed.plot(label=\"actual\")\n", "pred_series.plot(label=\"our RNN\")\n", "pred_series_ets.plot(label=\"ETS\")\n", "plt.legend()\n", "print(\"RNN MAPE:\", mape(pred_series, val_sp_transformed))\n", "print(\"ETS MAPE:\", mape(pred_series_ets, val_sp_transformed))" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "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": "3.12.11" }, "widgets": { "application/vnd.jupyter.widget-state+json": { "state": { "12af59d16c2a4424bddbf8b5143d6fc6": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "143d9b37c4074491b4e09b88774082d6": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": { "display": "inline-flex", "flex_flow": "row wrap", "width": "100%" } }, "165b7a19b5e1431a8178845e030f6fed": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "ProgressStyleModel", "state": { "description_width": "" } }, "2263354701ed421ab77493941c9758c0": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLModel", "state": { "layout": "IPY_MODEL_d0c0e073af3f401ebecd7f7bdd243b6e", "style": "IPY_MODEL_b5af0a32b84d49d6b36d6ff4158e4ff8", "value": " 7/7 [00:00<00:00, 122.89it/s, v_num=logs, train_loss=0.000237, val_loss=0.002]" } }, "2313d603f46d4b058e780bfe3d351e8e": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": { "display": "inline-flex", "flex_flow": "row wrap", "width": "100%" } }, "26bacdfa09394f7294ed5836ca4ca909": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLModel", "state": { "layout": "IPY_MODEL_7af15550cfe44494ba8594203ed0482f", "style": "IPY_MODEL_80f76d26decc456db377b095514c1ffb", "value": " 67/67 [00:04<00:00, 15.74it/s, v_num=logs, train_loss=0.0128, val_loss=0.040]" } }, "2c0571fbe59f4b92b9df611ee49ca12d": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "3599fa8c213b4016a188ad73fa6759ef": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "FloatProgressModel", "state": { "bar_style": "success", "layout": "IPY_MODEL_f4ade26d7c4c4e64b32229a4548f4a10", "max": 7, "style": "IPY_MODEL_4048e344d0944901a9e98881a2ba47a2", "value": 7 } }, "3adced8c017345bb8da26830859f3db9": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "FloatProgressModel", "state": { "bar_style": "success", "layout": "IPY_MODEL_12af59d16c2a4424bddbf8b5143d6fc6", "max": 49, "style": "IPY_MODEL_4c70b89cfc2d4ad2b85aafde6e033ffa", "value": 49 } }, "4048e344d0944901a9e98881a2ba47a2": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "ProgressStyleModel", "state": { "description_width": "" } }, "452eeadab82f4699ac5c3fd5ceaf18c6": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "4c70b89cfc2d4ad2b85aafde6e033ffa": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "ProgressStyleModel", "state": { "description_width": "" } }, "53b5a2adc68947c5a5b06ac12fac4abc": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "FloatProgressModel", "state": { "bar_style": "success", "layout": "IPY_MODEL_884abb83510a4480be5234e8a610c2bc", "max": 67, "style": "IPY_MODEL_165b7a19b5e1431a8178845e030f6fed", "value": 67 } }, "5b2e222d6ddb4b63aeaa1e520b04dae9": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HBoxModel", "state": { "children": [ "IPY_MODEL_7bb5acae0d1e4c7798ab444af7d9ef52", "IPY_MODEL_3adced8c017345bb8da26830859f3db9", "IPY_MODEL_fe0c3168efe1434eb01f1062d50e8020" ], "layout": "IPY_MODEL_2c0571fbe59f4b92b9df611ee49ca12d" } }, "5d68599a153c4affb8131b573ae709d0": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HBoxModel", "state": { "children": [ "IPY_MODEL_922cf01fb4334cf696bd4ed53c1ca495", "IPY_MODEL_53b5a2adc68947c5a5b06ac12fac4abc", "IPY_MODEL_26bacdfa09394f7294ed5836ca4ca909" ], "layout": "IPY_MODEL_2313d603f46d4b058e780bfe3d351e8e" } }, "7af15550cfe44494ba8594203ed0482f": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "7bb5acae0d1e4c7798ab444af7d9ef52": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLModel", "state": { "layout": "IPY_MODEL_ab2900975eaf4e9fb3ff9efb9f3025d9", "style": "IPY_MODEL_e020515573d3422b9bda0fa2e3f8e38d", "value": "historical forecasts: 100%" } }, "7e18d8e02b274b88955b6710a7ab90c2": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HBoxModel", "state": { "children": [ "IPY_MODEL_a826f3114fa94bf69d2583a199fcec13", "IPY_MODEL_3599fa8c213b4016a188ad73fa6759ef", "IPY_MODEL_2263354701ed421ab77493941c9758c0" ], "layout": "IPY_MODEL_143d9b37c4074491b4e09b88774082d6" } }, "80f76d26decc456db377b095514c1ffb": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLStyleModel", "state": { "description_width": "", "font_size": null, "text_color": null } }, "884abb83510a4480be5234e8a610c2bc": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": { "flex": "2" } }, "922cf01fb4334cf696bd4ed53c1ca495": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLModel", "state": { "layout": "IPY_MODEL_abc37802e2b94f6bbfad9255b0cfb60a", "style": "IPY_MODEL_c37adc29220e456aa08e3b3c23a257aa", "value": "Epoch 99: 100%" } }, "a826f3114fa94bf69d2583a199fcec13": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLModel", "state": { "layout": "IPY_MODEL_452eeadab82f4699ac5c3fd5ceaf18c6", "style": "IPY_MODEL_e6dcaa0400fb4e1d9ba2599baf5c73e9", "value": "Epoch 299: 100%" } }, "ab2900975eaf4e9fb3ff9efb9f3025d9": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "abc37802e2b94f6bbfad9255b0cfb60a": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "b5af0a32b84d49d6b36d6ff4158e4ff8": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLStyleModel", "state": { "description_width": "", "font_size": null, "text_color": null } }, "c37adc29220e456aa08e3b3c23a257aa": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLStyleModel", "state": { "description_width": "", "font_size": null, "text_color": null } }, "c553c42a891948648fa39f2f36b33bca": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "d0c0e073af3f401ebecd7f7bdd243b6e": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": {} }, "d3b7443372cb4833b8c32001843c8b13": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLStyleModel", "state": { "description_width": "", "font_size": null, "text_color": null } }, "e020515573d3422b9bda0fa2e3f8e38d": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLStyleModel", "state": { "description_width": "", "font_size": null, "text_color": null } }, "e6dcaa0400fb4e1d9ba2599baf5c73e9": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLStyleModel", "state": { "description_width": "", "font_size": null, "text_color": null } }, "f4ade26d7c4c4e64b32229a4548f4a10": { "model_module": "@jupyter-widgets/base", "model_module_version": "2.0.0", "model_name": "LayoutModel", "state": { "flex": "2" } }, "fe0c3168efe1434eb01f1062d50e8020": { "model_module": "@jupyter-widgets/controls", "model_module_version": "2.0.0", "model_name": "HTMLModel", "state": { "layout": "IPY_MODEL_c553c42a891948648fa39f2f36b33bca", "style": "IPY_MODEL_d3b7443372cb4833b8c32001843c8b13", "value": " 49/49 [00:29<00:00,  1.59it/s]" } } }, "version_major": 2, "version_minor": 0 } } }, "nbformat": 4, "nbformat_minor": 4 }