6.3.1 温度预测问题.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "数据集下载：\n",
    "\n",
    "cd ~/Downloads\n",
    "\n",
    "mkdir jena_climate\n",
    "\n",
    "cd jena_climate\n",
    "\n",
    "axel -c  https://s3.amazonaws.com/keras-datasets/jena_climate_2009_2016.csv.zip\n",
    "\n",
    "unzip jena_climate_2009_2016.csv.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['\"Date Time\"', '\"p (mbar)\"', '\"T (degC)\"', '\"Tpot (K)\"', '\"Tdew (degC)\"', '\"rh (%)\"', '\"VPmax (mbar)\"', '\"VPact (mbar)\"', '\"VPdef (mbar)\"', '\"sh (g/kg)\"', '\"H2OC (mmol/mol)\"', '\"rho (g/m**3)\"', '\"wv (m/s)\"', '\"max. wv (m/s)\"', '\"wd (deg)\"']\n",
      "420551\n"
     ]
    }
   ],
   "source": [
    "# 　观察耶拿天气数据集的数据\n",
    "import os\n",
    "data_dir = '/home/fc/Downloads/jena_climate'\n",
    "fname = os.path.join(data_dir, 'jena_climate_2009_2016.csv')\n",
    "f = open(fname)\n",
    "data = f.read()\n",
    "f.close()\n",
    "lines = data.split('\\n')\n",
    "header = lines[0].split(',')\n",
    "lines = lines[1:]\n",
    "print(header)\n",
    "print(len(lines))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 解析数据\n",
    "import numpy as np\n",
    "float_data = np.zeros((len(lines), len(header) - 1))\n",
    "for i, line in enumerate(lines):\n",
    "    values = [float(x) for x in line.split(',')[1:]]\n",
    "    float_data[i, :] = values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f4ca9eb5f28>]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4cb176ff60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 绘制温度时间序列\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "temp = float_data[:, 1] # 温度（单位：摄氏度）\n",
    "plt.plot(range(len(temp)), temp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f4ca9e35e48>]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmDyFDQDDMEyewgaAYRgmT/l/Xzad0b7V3m8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4ca9f407b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 绘制前 10 天的温度时间序列\n",
    "plt.plot(range(1440), temp[:1440])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#准备数据\n",
    "# 数据标准化\n",
    "mean = float_data[:200000].mean(axis=0)\n",
    "float_data -= mean\n",
    "std = float_data[:200000].std(axis=0)\n",
    "float_data /= std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 　生成时间序列样本及其目标的生成器\n",
    "def generator(data, lookback, delay, min_index, max_index,\n",
    "    shuffle=False, batch_size=128, step=6):\n",
    "    if max_index is None:\n",
    "        max_index = len(data) - delay - 1\n",
    "    i = min_index + lookback\n",
    "    while 1:\n",
    "        if shuffle:\n",
    "            rows = np.random.randint(\n",
    "            min_index + lookback, max_index, size=batch_size)\n",
    "        else:\n",
    "            if i + batch_size >= max_index:\n",
    "                i = min_index + lookback\n",
    "            rows = np.arange(i, min(i + batch_size, max_index))\n",
    "            i += len(rows)\n",
    "        samples = np.zeros((len(rows),\n",
    "                            lookback // step,\n",
    "                            data.shape[-1]))\n",
    "        targets = np.zeros((len(rows),))\n",
    "        for j, row in enumerate(rows):\n",
    "            indices = range(rows[j] - lookback, rows[j], step)\n",
    "            samples[j] = data[indices]\n",
    "            targets[j] = data[rows[j] + delay][1]\n",
    "        yield samples, targets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 准备训练生成器、验证生成器和测试生成器\n",
    "lookback = 1440\n",
    "step = 6\n",
    "delay = 144\n",
    "batch_size = 128\n",
    "train_gen = generator(float_data,\n",
    "                      lookback=lookback,\n",
    "                      delay=delay,\n",
    "                      min_index=0,\n",
    "                      max_index=200000,\n",
    "                      shuffle=True,\n",
    "                      step=step,\n",
    "                      batch_size=batch_size)\n",
    "val_gen = generator(float_data,\n",
    "                    lookback=lookback,\n",
    "                    delay=delay,\n",
    "                    min_index=200001,\n",
    "                    max_index=300000,\n",
    "                    step=step,\n",
    "                    batch_size=batch_size)\n",
    "test_gen = generator(float_data,\n",
    "                     lookback=lookback,\n",
    "                     delay=delay,\n",
    "                     min_index=300001,\n",
    "                     max_index=None,\n",
    "                     step=step,\n",
    "                     batch_size=batch_size)\n",
    "val_steps = (300000 - 200001 - lookback) //batch_size\n",
    "test_steps = (len(float_data) - 300001 - lookback) //batch_size "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.289735972991\n"
     ]
    }
   ],
   "source": [
    "# 　计算符合常识的基准方法的 MAE\n",
    "def evaluate_naive_method():\n",
    "    batch_maes = []\n",
    "    for step in range(val_steps):\n",
    "        samples, targets = next(val_gen)\n",
    "        preds = samples[:, -1, 1]\n",
    "        mae = np.mean(np.abs(preds - targets))\n",
    "        batch_maes.append(mae)\n",
    "    print(np.mean(batch_maes))\n",
    "evaluate_naive_method()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 将 MAE 转换成摄氏温度误差\n",
    "celsius_mae = 0.29 * std[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/fc/anaconda3/lib/python3.6/site-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n",
      "Epoch 1/20\n",
      "500/500 [==============================] - 14s 28ms/step - loss: 1.4025 - val_loss: 0.6556\n",
      "Epoch 2/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.4543 - val_loss: 0.3261\n",
      "Epoch 3/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2953 - val_loss: 0.3204\n",
      "Epoch 4/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2663 - val_loss: 0.3407\n",
      "Epoch 5/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2520 - val_loss: 0.3388\n",
      "Epoch 6/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2412 - val_loss: 0.3296\n",
      "Epoch 7/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2350 - val_loss: 0.3150\n",
      "Epoch 8/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2302 - val_loss: 0.3378\n",
      "Epoch 9/20\n",
      "500/500 [==============================] - 13s 26ms/step - loss: 0.2237 - val_loss: 0.3655\n",
      "Epoch 10/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2200 - val_loss: 0.3194\n",
      "Epoch 11/20\n",
      "500/500 [==============================] - 14s 28ms/step - loss: 0.2188 - val_loss: 0.3533\n",
      "Epoch 12/20\n",
      "500/500 [==============================] - 14s 28ms/step - loss: 0.2154 - val_loss: 0.3461\n",
      "Epoch 13/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2131 - val_loss: 0.3483\n",
      "Epoch 14/20\n",
      "500/500 [==============================] - 13s 26ms/step - loss: 0.2094 - val_loss: 0.3455\n",
      "Epoch 15/20\n",
      "500/500 [==============================] - 13s 26ms/step - loss: 0.2055 - val_loss: 0.3225\n",
      "Epoch 16/20\n",
      "500/500 [==============================] - 13s 26ms/step - loss: 0.2062 - val_loss: 0.3455\n",
      "Epoch 17/20\n",
      "500/500 [==============================] - 13s 26ms/step - loss: 0.2024 - val_loss: 0.3322\n",
      "Epoch 18/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.2026 - val_loss: 0.3288\n",
      "Epoch 19/20\n",
      "500/500 [==============================] - 13s 27ms/step - loss: 0.1996 - val_loss: 0.3580\n",
      "Epoch 20/20\n",
      "500/500 [==============================] - 13s 26ms/step - loss: 0.1968 - val_loss: 0.3327\n"
     ]
    }
   ],
   "source": [
    "# 　训练并评估一个密集连接模型\n",
    "from keras.models import Sequential\n",
    "from keras import layers\n",
    "from keras.optimizers import RMSprop\n",
    "model = Sequential()\n",
    "model.add(layers.Flatten(input_shape=(lookback // step, float_data.shape[-1])))\n",
    "model.add(layers.Dense(32, activation='relu'))\n",
    "model.add(layers.Dense(1))\n",
    "model.compile(optimizer=RMSprop(), loss='mae')\n",
    "history = model.fit_generator(train_gen,\n",
    " steps_per_epoch=500,\n",
    " epochs=20,\n",
    " validation_data=val_gen,\n",
    " validation_steps=val_steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFO4iIhH0/wFoEDGuMTtDxwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4c88d03978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 绘制结果\n",
    "import matplotlib.pyplot as plt\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "epochs = range(1, len(loss) + 1)\n",
    "plt.figure()\n",
    "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}