From a99cada83a8bb2e2013db19b6b326aef6740b272 Mon Sep 17 00:00:00 2001 From: Christiaan Meijer Date: Wed, 21 Dec 2022 15:59:14 +0100 Subject: [PATCH] add regression tutorial draft WIP, refs #297 --- notebooks/tutorial/regression_tutorial.ipynb | 295 +++++++++++++++++++ 1 file changed, 295 insertions(+) create mode 100644 notebooks/tutorial/regression_tutorial.ipynb diff --git a/notebooks/tutorial/regression_tutorial.ipynb b/notebooks/tutorial/regression_tutorial.ipynb new file mode 100644 index 0000000..a1bd730 --- /dev/null +++ b/notebooks/tutorial/regression_tutorial.ipynb @@ -0,0 +1,295 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "collapsed": true, + "pycharm": { + "name": "#%%\n" + } + }, + "outputs": [], + "source": [ + "import sys\n", + "import os\n", + "import numpy as np\n", + "import pandas as pd\n", + "\n", + "import mcfly\n", + "import tensorflow as tf\n", + "np.random.seed(2)" + ] + }, + { + "cell_type": "markdown", + "source": [ + "# Load data\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 5, + "outputs": [], + "source": [ + "from utils import tutorial_weather\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 6, + "outputs": [], + "source": [ + "X_train, X_test, y_train, y_test, map_path = tutorial_weather.load_data()\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 7, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "X_train.shape=(767, 89)\n", + "X_test.shape=(329, 89)\n", + "y_train.shape=(767,)\n", + "y_test.shape=(329,)\n" + ] + } + ], + "source": [ + "print(f'{X_train.shape=}')\n", + "print(f'{X_test.shape=}')\n", + "print(f'{y_train.shape=}')\n", + "print(f'{y_test.shape=}')" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 8, + "outputs": [ + { + "data": { + "text/plain": "count 767.000000\nmean 4.229074\nstd 3.853865\nmin 0.000000\n25% 0.900000\n50% 3.400000\n75% 6.700000\nmax 15.200000\nName: MAASTRICHT_sunshine, dtype: float64" + }, + "execution_count": 8, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "y_train.describe()\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "markdown", + "source": [ + "For Maastricht we have the following features in the dataset." + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 9, + "outputs": [ + { + "data": { + "text/plain": "['MAASTRICHT_cloud_cover',\n 'MAASTRICHT_humidity',\n 'MAASTRICHT_pressure',\n 'MAASTRICHT_global_radiation',\n 'MAASTRICHT_precipitation',\n 'MAASTRICHT_sunshine',\n 'MAASTRICHT_temp_mean',\n 'MAASTRICHT_temp_min',\n 'MAASTRICHT_temp_max']" + }, + "execution_count": 9, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "[e for e in X_train.columns if e[:5] == 'MAAST']" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "code", + "execution_count": null, + "outputs": [], + "source": [], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "markdown", + "source": [ + "The same features are available for all 18 locations below. We will use all features from all the locations to predict the sunshine in Maastricht on the next day.\n", + "![](https://zenodo.org/record/7053722/files/weather_prediction_dataset_map.jpg?download=1)\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + } + }, + { + "cell_type": "markdown", + "source": [ + "Let's have a look at the sunshine hours per day through the 3 years of data that we selected. The data is shuffled, so we need to sort it first to inspect any trends." + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 14, + "outputs": [ + { + "data": { + "text/plain": "" + }, + "execution_count": 14, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "text/plain": "
", + "image/png": "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\n" + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "y_train.sort_index().plot()\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 25, + "outputs": [ + { + "ename": "ValueError", + "evalue": "The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()", + "output_type": "error", + "traceback": [ + "\u001B[1;31m---------------------------------------------------------------------------\u001B[0m", + "\u001B[1;31mValueError\u001B[0m Traceback (most recent call last)", + "Cell \u001B[1;32mIn[25], line 2\u001B[0m\n\u001B[0;32m 1\u001B[0m metric \u001B[38;5;241m=\u001B[39m \u001B[38;5;124m'\u001B[39m\u001B[38;5;124mmean_squared_error\u001B[39m\u001B[38;5;124m'\u001B[39m\n\u001B[1;32m----> 2\u001B[0m models \u001B[38;5;241m=\u001B[39m \u001B[43mmcfly\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mmodelgen\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mgenerate_models\u001B[49m\u001B[43m(\u001B[49m\u001B[43mnp\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mexpand_dims\u001B[49m\u001B[43m(\u001B[49m\u001B[43mX_train\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mto_numpy\u001B[49m\u001B[43m(\u001B[49m\u001B[43m)\u001B[49m\u001B[43m,\u001B[49m\u001B[43m \u001B[49m\u001B[38;5;241;43m2\u001B[39;49m\u001B[43m)\u001B[49m\u001B[43m,\u001B[49m\n\u001B[0;32m 3\u001B[0m \u001B[43m \u001B[49m\u001B[43mnumber_of_classes\u001B[49m\u001B[38;5;241;43m=\u001B[39;49m\u001B[43m \u001B[49m\u001B[38;5;241;43m1\u001B[39;49m\u001B[43m,\u001B[49m\n\u001B[0;32m 4\u001B[0m \u001B[43m \u001B[49m\u001B[43mnumber_of_models\u001B[49m\u001B[43m \u001B[49m\u001B[38;5;241;43m=\u001B[39;49m\u001B[43m \u001B[49m\u001B[38;5;241;43m8\u001B[39;49m\u001B[43m,\u001B[49m\n\u001B[0;32m 5\u001B[0m \u001B[43m \u001B[49m\u001B[43mmetrics\u001B[49m\u001B[38;5;241;43m=\u001B[39;49m\u001B[43m[\u001B[49m\u001B[43mmetric\u001B[49m\u001B[43m]\u001B[49m\u001B[43m)\u001B[49m\n", + "File \u001B[1;32mc:\\users\\christiaanmeijer\\documents\\mcfly\\mcfly\\modelgen.py:111\u001B[0m, in \u001B[0;36mgenerate_models\u001B[1;34m(x_shape, number_of_classes, number_of_models, model_types, metrics, **hyperparameter_ranges)\u001B[0m\n\u001B[0;32m 108\u001B[0m np.random.shuffle(model_types)\n\u001B[0;32m 109\u001B[0m model_types_selected.extend(model_types)\n\u001B[1;32m--> 111\u001B[0m # Create list of Keras models and their hyperparameters\n\u001B[0;32m 112\u001B[0m # -------------------------------------------------------------------------\n\u001B[0;32m 113\u001B[0m models = []\n\u001B[0;32m 114\u001B[0m for current_model_type in model_types_selected[:number_of_models]:\n", + "File \u001B[1;32mc:\\users\\christiaanmeijer\\documents\\mcfly\\mcfly\\models\\inception_time.py:55\u001B[0m, in \u001B[0;36m__init__\u001B[1;34m(self, x_shape, number_of_classes, metrics, IT_min_network_depth, IT_max_network_depth, IT_min_filters_number, IT_max_filters_number, IT_min_max_kernel_size, IT_max_max_kernel_size, **_other)\u001B[0m\n\u001B[0;32m 52\u001B[0m self.number_of_classes = number_of_classes\n\u001B[0;32m 53\u001B[0m self.metrics = metrics\n\u001B[1;32m---> 55\u001B[0m # Limit parameter space based on input\n\u001B[0;32m 56\u001B[0m if IT_max_max_kernel_size > self.x_shape[1]:\n\u001B[0;32m 57\u001B[0m print(\"Set maximum kernel size for InceptionTime models to number of timesteps.\")\n", + "\u001B[1;31mValueError\u001B[0m: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()" + ] + } + ], + "source": [ + "metric = 'mean_squared_error'\n", + "models = mcfly.modelgen.generate_models(np.expand_dims(X_train.to_numpy(), 2),\n", + " number_of_classes= 1,\n", + " number_of_models = 8,\n", + " metrics=[metric])\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + }, + { + "cell_type": "code", + "execution_count": 24, + "outputs": [ + { + "data": { + "text/plain": "array([[[ 7. ],\n [ 0.73 ],\n [ 1.0194],\n ...,\n [13.1 ],\n [ 9.9 ],\n [16.3 ]],\n\n [[ 2. ],\n [ 0.54 ],\n [ 1.0154],\n ...,\n [19.9 ],\n [13.9 ],\n [25.9 ]],\n\n [[ 7. ],\n [ 0.91 ],\n [ 0.9929],\n ...,\n [ 4. ],\n [ 1.9 ],\n [ 6. ]],\n\n ...,\n\n [[ 7. ],\n [ 0.79 ],\n [ 1.0163],\n ...,\n [ 9.4 ],\n [ 5.1 ],\n [13.7 ]],\n\n [[ 7. ],\n [ 0.82 ],\n [ 1.0134],\n ...,\n [16.8 ],\n [13. ],\n [20.7 ]],\n\n [[ 0. ],\n [ 0.81 ],\n [ 1.0341],\n ...,\n [ 3.6 ],\n [ 0.9 ],\n [ 6.3 ]]])" + }, + "execution_count": 24, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "\n" + ], + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%%\n" + } + } + } + ], + "metadata": { + "kernelspec": { + "name": "mcfly", + "language": "python", + "display_name": "mcfly" + }, + "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": 0 +} \ No newline at end of file