{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "<img width=\"800px\" src=\"../fidle/img/00-Fidle-header-01.svg\"></img>\n", "\n", "# <!-- TITLE --> [SYNOP2] - Time series with RNN - Try a prediction\n", "<!-- DESC --> Episode 2 : Training session and first predictions\n", "<!-- AUTHOR : Jean-Luc Parouty (CNRS/SIMaP) -->\n", "\n", "## Objectives :\n", " - Make a simple prediction (3h)\n", " - Understanding the use of a recurrent neural network\n", "\n", "\n", "SYNOP meteorological data, available at: https://public.opendatasoft.com\n", "\n", "## What we're going to do :\n", "\n", " - Read our dataset\n", " - Select our data and normalize it\n", " - Doing our training\n", " - Making simple predictions\n", "\n", "## Step 1 - Import and init\n", "### 1.1 - Python" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "data": { "text/html": [ "<style>\n", "\n", "div.warn { \n", " background-color: #fcf2f2;\n", " border-color: #dFb5b4;\n", " border-left: 5px solid #dfb5b4;\n", " padding: 0.5em;\n", " font-weight: bold;\n", " font-size: 1.1em;;\n", " }\n", "\n", "\n", "\n", "div.nota { \n", " background-color: #DAFFDE;\n", " border-left: 5px solid #92CC99;\n", " padding: 0.5em;\n", " }\n", "\n", "div.todo:before { content:url(data:image/svg+xml;base64,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);\n", " float:left;\n", " margin-right:20px;\n", " margin-top:-20px;\n", " margin-bottom:20px;\n", "}\n", "div.todo{\n", " font-weight: bold;\n", " font-size: 1.1em;\n", " margin-top:40px;\n", "}\n", "div.todo ul{\n", " margin: 0.2em;\n", "}\n", "div.todo li{\n", " margin-left:60px;\n", " margin-top:0;\n", " margin-bottom:0;\n", "}\n", "\n", "\n", "</style>\n", "\n" ], "text/plain": [ "<IPython.core.display.HTML object>" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "FIDLE 2020 - Practical Work Module\n", "Version : 0.57 DEV\n", "Run time : Thursday 10 September 2020, 17:06:53\n", "TensorFlow version : 2.2.0\n", "Keras version : 2.3.0-tf\n", "Current place : Fidle at IDRIS\n", "Dataset dir : /gpfswork/rech/mlh/commun/datasets\n", "Update keras cache : Done\n" ] } ], "source": [ "import tensorflow as tf\n", "from tensorflow import keras\n", "from tensorflow.keras.callbacks import TensorBoard\n", "from tensorflow.keras.preprocessing.sequence import TimeseriesGenerator\n", "\n", "import numpy as np\n", "import math, random\n", "import matplotlib.pyplot as plt\n", "\n", "import pandas as pd\n", "import h5py, json\n", "import os,time,sys\n", "\n", "from importlib import reload\n", "\n", "sys.path.append('..')\n", "import fidle.pwk as ooo\n", "\n", "place, datasets_dir = ooo.init()\n", "\n", "def np_print(*args):\n", " with np.printoptions(formatter={'float':'{:8.2f}'.format}, linewidth=np.inf):\n", " for a in args:\n", " print(a) " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 2 - Read and prepare dataset\n", "### 2.1 - Read it" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "data": { "text/markdown": [ "<br>**Train dataset example :**" ], "text/plain": [ "<IPython.core.display.Markdown object>" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "<div>\n", "<style scoped>\n", " .dataframe tbody tr th:only-of-type {\n", " vertical-align: middle;\n", " }\n", "\n", " .dataframe tbody tr th {\n", " vertical-align: top;\n", " }\n", "\n", " .dataframe thead th {\n", " text-align: right;\n", " }\n", "</style>\n", "<table border=\"1\" class=\"dataframe\">\n", " <thead>\n", " <tr style=\"text-align: right;\">\n", " <th></th>\n", " <th>tend</th>\n", " <th>cod_tend</th>\n", " <th>dd</th>\n", " <th>ff</th>\n", " <th>td</th>\n", " <th>u</th>\n", " <th>ww</th>\n", " <th>pres</th>\n", " <th>rafper</th>\n", " <th>rr1</th>\n", " <th>rr3</th>\n", " <th>tc</th>\n", " </tr>\n", " </thead>\n", " <tbody>\n", " <tr>\n", " <th>0</th>\n", " <td>-120.0</td>\n", " <td>6.0</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>278.75</td>\n", " <td>88.0</td>\n", " <td>60.0</td>\n", " <td>96250.0</td>\n", " <td>4.1</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>7.5</td>\n", " </tr>\n", " <tr>\n", " <th>1</th>\n", " <td>-150.0</td>\n", " <td>6.0</td>\n", " <td>60.0</td>\n", " <td>1.0</td>\n", " <td>278.65</td>\n", " <td>93.0</td>\n", " <td>61.0</td>\n", " <td>96100.0</td>\n", " <td>2.6</td>\n", " <td>0.2</td>\n", " <td>0.6</td>\n", " <td>6.6</td>\n", " </tr>\n", " <tr>\n", " <th>2</th>\n", " <td>10.0</td>\n", " <td>3.0</td>\n", " <td>280.0</td>\n", " <td>2.1</td>\n", " <td>278.85</td>\n", " <td>95.0</td>\n", " <td>58.0</td>\n", " <td>96110.0</td>\n", " <td>2.6</td>\n", " <td>0.0</td>\n", " <td>0.4</td>\n", " <td>6.4</td>\n", " </tr>\n", " <tr>\n", " <th>3</th>\n", " <td>230.0</td>\n", " <td>3.0</td>\n", " <td>310.0</td>\n", " <td>2.6</td>\n", " <td>279.15</td>\n", " <td>96.0</td>\n", " <td>50.0</td>\n", " <td>96340.0</td>\n", " <td>5.7</td>\n", " <td>0.0</td>\n", " <td>3.0</td>\n", " <td>6.6</td>\n", " </tr>\n", " <tr>\n", " <th>4</th>\n", " <td>280.0</td>\n", " <td>1.0</td>\n", " <td>330.0</td>\n", " <td>4.6</td>\n", " <td>278.15</td>\n", " <td>94.0</td>\n", " <td>21.0</td>\n", " <td>96620.0</td>\n", " <td>8.7</td>\n", " <td>0.4</td>\n", " <td>0.8</td>\n", " <td>5.9</td>\n", " </tr>\n", " <tr>\n", " <th>5</th>\n", " <td>480.0</td>\n", " <td>3.0</td>\n", " <td>350.0</td>\n", " <td>5.1</td>\n", " <td>276.95</td>\n", " <td>91.0</td>\n", " <td>60.0</td>\n", " <td>97100.0</td>\n", " <td>8.2</td>\n", " <td>0.2</td>\n", " <td>0.4</td>\n", " <td>5.2</td>\n", " </tr>\n", " <tr>\n", " <th>6</th>\n", " <td>530.0</td>\n", " <td>2.0</td>\n", " <td>350.0</td>\n", " <td>3.1</td>\n", " <td>274.05</td>\n", " <td>83.0</td>\n", " <td>21.0</td>\n", " <td>97630.0</td>\n", " <td>7.2</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>3.5</td>\n", " </tr>\n", " <tr>\n", " <th>7</th>\n", " <td>450.0</td>\n", " <td>2.0</td>\n", " <td>340.0</td>\n", " <td>6.2</td>\n", " <td>272.15</td>\n", " <td>81.0</td>\n", " <td>2.0</td>\n", " <td>98080.0</td>\n", " <td>9.3</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>1.9</td>\n", " </tr>\n", " <tr>\n", " <th>8</th>\n", " <td>280.0</td>\n", " <td>1.0</td>\n", " <td>320.0</td>\n", " <td>6.2</td>\n", " <td>270.15</td>\n", " <td>74.0</td>\n", " <td>2.0</td>\n", " <td>98360.0</td>\n", " <td>10.3</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>1.1</td>\n", " </tr>\n", " <tr>\n", " <th>9</th>\n", " <td>220.0</td>\n", " <td>1.0</td>\n", " <td>290.0</td>\n", " <td>2.6</td>\n", " <td>269.65</td>\n", " <td>72.0</td>\n", " <td>2.0</td>\n", " <td>98580.0</td>\n", " <td>5.1</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>1.0</td>\n", " </tr>\n", " <tr>\n", " <th>10</th>\n", " <td>100.0</td>\n", " <td>1.0</td>\n", " <td>350.0</td>\n", " <td>3.1</td>\n", " <td>270.45</td>\n", " <td>79.0</td>\n", " <td>2.0</td>\n", " <td>98680.0</td>\n", " <td>4.1</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>0.5</td>\n", " </tr>\n", " <tr>\n", " <th>11</th>\n", " <td>300.0</td>\n", " <td>3.0</td>\n", " <td>350.0</td>\n", " <td>5.1</td>\n", " <td>268.55</td>\n", " <td>70.0</td>\n", " <td>2.0</td>\n", " <td>98980.0</td>\n", " <td>6.7</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>-0.3</td>\n", " </tr>\n", " <tr>\n", " <th>12</th>\n", " <td>130.0</td>\n", " <td>1.0</td>\n", " <td>10.0</td>\n", " <td>4.6</td>\n", " <td>267.45</td>\n", " <td>60.0</td>\n", " <td>2.0</td>\n", " <td>99110.0</td>\n", " <td>7.7</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>1.2</td>\n", " </tr>\n", " <tr>\n", " <th>13</th>\n", " <td>150.0</td>\n", " <td>3.0</td>\n", " <td>10.0</td>\n", " <td>5.7</td>\n", " <td>267.45</td>\n", " <td>59.0</td>\n", " <td>2.0</td>\n", " <td>99260.0</td>\n", " <td>8.7</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>1.5</td>\n", " </tr>\n", " <tr>\n", " <th>14</th>\n", " <td>140.0</td>\n", " <td>1.0</td>\n", " <td>50.0</td>\n", " <td>2.6</td>\n", " <td>268.15</td>\n", " <td>70.0</td>\n", " <td>2.0</td>\n", " <td>99400.0</td>\n", " <td>5.7</td>\n", " <td>0.0</td>\n", " <td>0.0</td>\n", " <td>-0.8</td>\n", " </tr>\n", " </tbody>\n", "</table>\n", "</div>" ], "text/plain": [ " tend cod_tend dd ff td u ww pres rafper rr1 \\\n", "0 -120.0 6.0 0.0 0.0 278.75 88.0 60.0 96250.0 4.1 0.0 \n", "1 -150.0 6.0 60.0 1.0 278.65 93.0 61.0 96100.0 2.6 0.2 \n", "2 10.0 3.0 280.0 2.1 278.85 95.0 58.0 96110.0 2.6 0.0 \n", "3 230.0 3.0 310.0 2.6 279.15 96.0 50.0 96340.0 5.7 0.0 \n", "4 280.0 1.0 330.0 4.6 278.15 94.0 21.0 96620.0 8.7 0.4 \n", "5 480.0 3.0 350.0 5.1 276.95 91.0 60.0 97100.0 8.2 0.2 \n", "6 530.0 2.0 350.0 3.1 274.05 83.0 21.0 97630.0 7.2 0.0 \n", "7 450.0 2.0 340.0 6.2 272.15 81.0 2.0 98080.0 9.3 0.0 \n", "8 280.0 1.0 320.0 6.2 270.15 74.0 2.0 98360.0 10.3 0.0 \n", "9 220.0 1.0 290.0 2.6 269.65 72.0 2.0 98580.0 5.1 0.0 \n", "10 100.0 1.0 350.0 3.1 270.45 79.0 2.0 98680.0 4.1 0.0 \n", "11 300.0 3.0 350.0 5.1 268.55 70.0 2.0 98980.0 6.7 0.0 \n", "12 130.0 1.0 10.0 4.6 267.45 60.0 2.0 99110.0 7.7 0.0 \n", "13 150.0 3.0 10.0 5.7 267.45 59.0 2.0 99260.0 8.7 0.0 \n", "14 140.0 1.0 50.0 2.6 268.15 70.0 2.0 99400.0 5.7 0.0 \n", "\n", " rr3 tc \n", "0 0.0 7.5 \n", "1 0.6 6.6 \n", "2 0.4 6.4 \n", "3 3.0 6.6 \n", "4 0.8 5.9 \n", "5 0.4 5.2 \n", "6 0.0 3.5 \n", "7 0.0 1.9 \n", "8 0.0 1.1 \n", "9 0.0 1.0 \n", "10 0.0 0.5 \n", "11 0.0 -0.3 \n", "12 0.0 1.2 \n", "13 0.0 1.5 \n", "14 0.0 -0.8 " ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/markdown": [ "<br>**After normalization :**" ], "text/plain": [ "<IPython.core.display.Markdown object>" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "<style type=\"text/css\" >\n", "</style><table id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729\" ><thead> <tr> <th class=\"blank level0\" ></th> <th class=\"col_heading level0 col0\" >tend</th> <th class=\"col_heading level0 col1\" >cod_tend</th> <th class=\"col_heading level0 col2\" >dd</th> <th class=\"col_heading level0 col3\" >ff</th> <th class=\"col_heading level0 col4\" >td</th> <th class=\"col_heading level0 col5\" >u</th> <th class=\"col_heading level0 col6\" >ww</th> <th class=\"col_heading level0 col7\" >pres</th> <th class=\"col_heading level0 col8\" >rafper</th> <th class=\"col_heading level0 col9\" >rr1</th> <th class=\"col_heading level0 col10\" >rr3</th> <th class=\"col_heading level0 col11\" >tc</th> </tr></thead><tbody>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row0\" class=\"row_heading level0 row0\" >count</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col0\" class=\"data row0 col0\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col1\" class=\"data row0 col1\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col2\" class=\"data row0 col2\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col3\" class=\"data row0 col3\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col4\" class=\"data row0 col4\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col5\" class=\"data row0 col5\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col6\" class=\"data row0 col6\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col7\" class=\"data row0 col7\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col8\" class=\"data row0 col8\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col9\" class=\"data row0 col9\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col10\" class=\"data row0 col10\" >25000.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row0_col11\" class=\"data row0 col11\" >25000.00</td>\n", " </tr>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row1\" class=\"row_heading level0 row1\" >mean</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col0\" class=\"data row1 col0\" >-0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col1\" class=\"data row1 col1\" >-0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col2\" class=\"data row1 col2\" >0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col3\" class=\"data row1 col3\" >0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col4\" class=\"data row1 col4\" >-0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col5\" class=\"data row1 col5\" >0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col6\" class=\"data row1 col6\" >0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col7\" class=\"data row1 col7\" >-0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col8\" class=\"data row1 col8\" >-0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col9\" class=\"data row1 col9\" >-0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col10\" class=\"data row1 col10\" >0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row1_col11\" class=\"data row1 col11\" >-0.00</td>\n", " </tr>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row2\" class=\"row_heading level0 row2\" >std</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col0\" class=\"data row2 col0\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col1\" class=\"data row2 col1\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col2\" class=\"data row2 col2\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col3\" class=\"data row2 col3\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col4\" class=\"data row2 col4\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col5\" class=\"data row2 col5\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col6\" class=\"data row2 col6\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col7\" class=\"data row2 col7\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col8\" class=\"data row2 col8\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col9\" class=\"data row2 col9\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col10\" class=\"data row2 col10\" >1.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row2_col11\" class=\"data row2 col11\" >1.00</td>\n", " </tr>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row3\" class=\"row_heading level0 row3\" >min</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col0\" class=\"data row3 col0\" >-6.80</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col1\" class=\"data row3 col1\" >-1.59</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col2\" class=\"data row3 col2\" >-1.75</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col3\" class=\"data row3 col3\" >-1.37</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col4\" class=\"data row3 col4\" >-5.18</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col5\" class=\"data row3 col5\" >-3.82</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col6\" class=\"data row3 col6\" >-0.52</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col7\" class=\"data row3 col7\" >-4.94</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col8\" class=\"data row3 col8\" >-1.64</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col9\" class=\"data row3 col9\" >-0.31</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col10\" class=\"data row3 col10\" >-0.27</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row3_col11\" class=\"data row3 col11\" >-3.03</td>\n", " </tr>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row4\" class=\"row_heading level0 row4\" >25%</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col0\" class=\"data row4 col0\" >-0.64</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col1\" class=\"data row4 col1\" >-0.85</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col2\" class=\"data row4 col2\" >-0.64</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col3\" class=\"data row4 col3\" >-0.76</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col4\" class=\"data row4 col4\" >-0.72</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col5\" class=\"data row4 col5\" >-0.71</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col6\" class=\"data row4 col6\" >-0.42</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col7\" class=\"data row4 col7\" >-0.55</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col8\" class=\"data row4 col8\" >-0.69</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col9\" class=\"data row4 col9\" >-0.15</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col10\" class=\"data row4 col10\" >-0.20</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row4_col11\" class=\"data row4 col11\" >-0.75</td>\n", " </tr>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row5\" class=\"row_heading level0 row5\" >50%</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col0\" class=\"data row5 col0\" >-0.00</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col1\" class=\"data row5 col1\" >-0.48</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col2\" class=\"data row5 col2\" >-0.12</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col3\" class=\"data row5 col3\" >-0.19</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col4\" class=\"data row5 col4\" >0.05</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col5\" class=\"data row5 col5\" >0.18</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col6\" class=\"data row5 col6\" >-0.42</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col7\" class=\"data row5 col7\" >0.03</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col8\" class=\"data row5 col8\" >-0.27</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col9\" class=\"data row5 col9\" >-0.15</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col10\" class=\"data row5 col10\" >-0.20</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row5_col11\" class=\"data row5 col11\" >-0.01</td>\n", " </tr>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row6\" class=\"row_heading level0 row6\" >75%</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col0\" class=\"data row6 col0\" >0.63</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col1\" class=\"data row6 col1\" >0.99</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col2\" class=\"data row6 col2\" >1.08</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col3\" class=\"data row6 col3\" >0.50</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col4\" class=\"data row6 col4\" >0.79</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col5\" class=\"data row6 col5\" >0.84</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col6\" class=\"data row6 col6\" >-0.37</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col7\" class=\"data row6 col7\" >0.61</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col8\" class=\"data row6 col8\" >0.52</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col9\" class=\"data row6 col9\" >-0.15</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col10\" class=\"data row6 col10\" >-0.20</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row6_col11\" class=\"data row6 col11\" >0.72</td>\n", " </tr>\n", " <tr>\n", " <th id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729level0_row7\" class=\"row_heading level0 row7\" >max</th>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col0\" class=\"data row7 col0\" >7.16</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col1\" class=\"data row7 col1\" >1.36</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col2\" class=\"data row7 col2\" >1.34</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col3\" class=\"data row7 col3\" >6.28</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col4\" class=\"data row7 col4\" >2.40</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col5\" class=\"data row7 col5\" >1.62</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col6\" class=\"data row7 col6\" >4.46</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col7\" class=\"data row7 col7\" >3.10</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col8\" class=\"data row7 col8\" >6.29</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col9\" class=\"data row7 col9\" >30.36</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col10\" class=\"data row7 col10\" >31.27</td>\n", " <td id=\"T_567f8ed0_f377_11ea_820e_0cc47af5c729row7_col11\" class=\"data row7 col11\" >3.02</td>\n", " </tr>\n", " </tbody></table>" ], "text/plain": [ "<pandas.io.formats.style.Styler at 0x1517638e7e50>" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "Dataset : (29165, 14)\n", "Train dataset : (25000, 12)\n", "Test dataset : (4165, 12)\n" ] } ], "source": [ "dataset_filename = 'synop-LYS.csv'\n", "schema_filename = 'synop.json'\n", "train_len = 25000\n", "features = ['tend', 'cod_tend', 'dd', 'ff', 'td', 'u', 'ww', 'pres', 'rafper', 'rr1', 'rr3', 'tc']\n", "features_len = len(features)\n", "\n", "# ---- Read dataset\n", "\n", "df = pd.read_csv(f'{datasets_dir}/SYNOP/{dataset_filename}', header=0, sep=';')\n", "\n", "# ---- Train / Test\n", "\n", "dataset_train = df.loc[ :train_len-1, features ]\n", "dataset_test = df.loc[train_len:, features ]\n", "ooo.subtitle('Train dataset example :')\n", "display(dataset_train.head(15))\n", "\n", "# ---- Normalize, and convert to numpy array\n", "\n", "mean = dataset_train.mean()\n", "std = dataset_train.std()\n", "dataset_train = (dataset_train - mean) / std\n", "dataset_test = (dataset_test - mean) / std\n", "\n", "ooo.subtitle('After normalization :')\n", "display(dataset_train.describe().style.format(\"{0:.2f}\"))\n", "\n", "dataset_train = dataset_train.to_numpy()\n", "dataset_test = dataset_test.to_numpy()\n", "\n", "\n", "print('Dataset : ',df.shape)\n", "print('Train dataset : ',dataset_train.shape)\n", "print('Test dataset : ',dataset_test.shape)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 2.2 - Prepare data generator" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Nombre de train batchs disponibles : 781\n", "batch x shape : (32, 16, 12)\n", "batch y shape : (32, 12)\n" ] } ], "source": [ "sequence_len = 16\n", "batch_size = 32\n", "\n", "# ---- Train generator\n", "train_generator = TimeseriesGenerator(dataset_train, dataset_train, length=sequence_len, batch_size=batch_size)\n", "test_generator = TimeseriesGenerator(dataset_test, dataset_test, length=sequence_len, batch_size=batch_size)\n", "\n", "# ---- About\n", "\n", "x,y=train_generator[0]\n", "print(f'Nombre de train batchs disponibles : ', len(train_generator))\n", "print('batch x shape : ',x.shape)\n", "print('batch y shape : ',y.shape)\n", "\n", "# x,y=train_generator[0]\n", "# np_print(x[0])\n", "# np_print(y[0])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 3 - Create a model" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "WARNING:tensorflow:Layer lstm will not use cuDNN kernel since it doesn't meet the cuDNN kernel criteria. It will use generic GPU kernel as fallback when running on GPU\n", "Model: \"sequential\"\n", "_________________________________________________________________\n", "Layer (type) Output Shape Param # \n", "=================================================================\n", "lstm (LSTM) (None, 100) 45200 \n", "_________________________________________________________________\n", "dropout (Dropout) (None, 100) 0 \n", "_________________________________________________________________\n", "dense (Dense) (None, 12) 1212 \n", "=================================================================\n", "Total params: 46,412\n", "Trainable params: 46,412\n", "Non-trainable params: 0\n", "_________________________________________________________________\n" ] } ], "source": [ "model = keras.models.Sequential()\n", "model.add( keras.layers.InputLayer(input_shape=(sequence_len, features_len)) )\n", "model.add( keras.layers.LSTM(100, activation='relu') )\n", "model.add( keras.layers.Dropout(0.2) )\n", "model.add( keras.layers.Dense(features_len) )\n", "\n", "model.summary()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Step 4 - Compile and run" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4.1 - Callback" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "os.makedirs('./run/models', mode=0o750, exist_ok=True)\n", "save_dir = \"./run/models/best_model.h5\"\n", "bestmodel_callback = tf.keras.callbacks.ModelCheckpoint(filepath=save_dir, verbose=0, save_best_only=True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4.2 - Compile" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "model.compile(optimizer='adam', \n", " loss='mse', \n", " metrics = ['mae'] )" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4.3 - Fit" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "WARNING:tensorflow:From <timed exec>:5: Model.fit_generator (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n", "Instructions for updating:\n", "Please use Model.fit, which supports generators.\n", "Epoch 1/10\n", "781/781 [==============================] - 14s 17ms/step - loss: 0.6023 - mae: 0.5071 - val_loss: 0.4669 - val_mae: 0.4289\n", "Epoch 2/10\n", "781/781 [==============================] - 13s 17ms/step - loss: 0.5045 - mae: 0.4379 - val_loss: 0.4392 - val_mae: 0.4006\n", "Epoch 3/10\n", "781/781 [==============================] - 13s 17ms/step - loss: 0.4809 - mae: 0.4192 - val_loss: 0.4367 - val_mae: 0.4023\n", "Epoch 4/10\n", "781/781 [==============================] - 13s 16ms/step - loss: 0.4703 - mae: 0.4090 - val_loss: 0.4305 - val_mae: 0.4091\n", "Epoch 5/10\n", "781/781 [==============================] - 13s 17ms/step - loss: 0.4574 - mae: 0.4019 - val_loss: 0.4155 - val_mae: 0.3849\n", "Epoch 6/10\n", "781/781 [==============================] - 13s 17ms/step - loss: 0.4503 - mae: 0.3954 - val_loss: 0.4062 - val_mae: 0.3730\n", "Epoch 7/10\n", "781/781 [==============================] - 13s 17ms/step - loss: 0.4440 - mae: 0.3916 - val_loss: 0.4108 - val_mae: 0.3709\n", "Epoch 8/10\n", "781/781 [==============================] - 13s 17ms/step - loss: 0.4390 - mae: 0.3882 - val_loss: 0.4200 - val_mae: 0.3758\n", "Epoch 9/10\n", "781/781 [==============================] - 13s 17ms/step - loss: 0.4343 - mae: 0.3866 - val_loss: 0.4074 - val_mae: 0.3673\n", "Epoch 10/10\n", "781/781 [==============================] - 14s 18ms/step - loss: 0.4282 - mae: 0.3838 - val_loss: 0.4044 - val_mae: 0.3660\n", "CPU times: user 4min 41s, sys: 21.2 s, total: 5min 2s\n", "Wall time: 2min 15s\n" ] } ], "source": [ "%%time\n", "\n", "history=model.fit_generator(train_generator, \n", " epochs=10, \n", " verbose=1,\n", " validation_data = test_generator,\n", " callbacks = [bestmodel_callback])" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 576x432 with 1 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 576x432 with 1 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "ooo.plot_history(history,plot={'loss':['loss','val_loss'], 'mae':['mae','val_mae']})" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 5 - Predict" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 5.1 - Load model" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "WARNING:tensorflow:Layer lstm will not use cuDNN kernel since it doesn't meet the cuDNN kernel criteria. It will use generic GPU kernel as fallback when running on GPU\n" ] } ], "source": [ "loaded_model = tf.keras.models.load_model('./run/models/best_model.h5')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 5.2 Make a prediction" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 1080x1152 with 12 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "s=random.randint(0,len(dataset_test)-sequence_len)\n", "\n", "sequence = dataset_test[s:s+sequence_len]\n", "sequence_true = dataset_test[s:s+sequence_len+1]\n", "\n", "pred = loaded_model.predict( np.array([sequence]) )\n", "\n", "# ---- Show result\n", "reload(ooo)\n", "ooo.plot_multivariate_serie(sequence_true, predictions=pred, labels=features)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 5.3 Full prediction" ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Gap between prediction and reality : 0.07 °C\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 3024x2304 with 1 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "def denormalize(mean,std,seq):\n", " nseq = seq.copy()\n", " for i,s in enumerate(nseq):\n", " s = s*std + mean\n", " nseq[i]=s\n", " return nseq\n", "\n", "\n", "# ---- Get a sequence\n", "\n", "i=random.randint(0,len(dataset_test)-sequence_len)\n", "sequence = dataset_test[i:i+sequence_len]\n", "sequence_true = dataset_test[i:i+sequence_len+1]\n", "\n", "# ---- Prediction\n", "\n", "pred = loaded_model.predict( np.array([sequence]) )\n", "\n", "# ---- De-normalization\n", "\n", "sequence_true = denormalize(mean,std, sequence_true)\n", "pred = denormalize(mean,std, pred)\n", "\n", "# ---- Show it\n", "feat=11\n", "\n", "delta_deg=abs(sequence_true[-1][feat]-pred[-1][feat])\n", "print(f'Gap between prediction and reality : {delta_deg:.2f} °C')\n", "\n", "reload(ooo)\n", "ooo.plot_multivariate_serie(sequence_true, predictions=pred, labels=features, only_features=[feat],width=14, height=8)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "---\n", "<img width=\"80px\" src=\"../fidle/img/00-Fidle-logo-01.svg\"></img>" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "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": "3.7.7" } }, "nbformat": 4, "nbformat_minor": 4 }