打开股票量化的黑箱-自己动手实现一个印钞机3.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 打开股票量化的黑箱(自己动手实现一个‘印钞机’) 第三章\n",
    "\n",
    "### 作者：阿布🐶\n",
    "\n",
    "### 未经本人允许禁止转载"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  ‘非均衡胜负收益’带来的必然’非均衡胜负比例‘，目标由’因子‘的能力解决一部分，’模式识别‘提升关键的一部分"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "本章开始说文章的核心了，**模式识别提升关键的一部分** 本章的内容主要是通过机器学习如svm，随机森林等对股票模式识别的初步探索，俗称罪恶的第一步，但还是要坚定的卖出，毕竟目标是印钞机"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 下面的工作运行因子对多年数据进行回测，模式识别中基本的需求就是生成训练集数据与测试集数据，对训练集的数据抽取特质，总结规律，在测试集上指导交易，与没有指导交易的测试集进行比对，查看效果"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import ZEnv\n",
    "import ZLog\n",
    "import ZCommonUtil\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "import BuyGoldenFactor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "BuyGoldenFactor.g_enable_filter_ml = True # 回测因子的历史且结果集加入机器学习需要的数据\n",
    "BuyGoldenFactor.g_enable_snap = True # 生成交易这一时刻60日图像，为深度学习准备, 详情见第四章\n",
    "buy_factors = [{'XD': 42, 'class': BuyGoldenFactorClass, 'draw': True}]\n",
    "\n",
    "# n_folds： 6年数据回测生成五年单子\n",
    "# train_test_split：回测时预留一份股票代码表作为测试集，生成9份训练集，当use_last_test为True 时只用一份测试股票代码表进行因子回测\n",
    "# ret_cnt_need ＝ 0：回测所有股票\n",
    "out, orders_pd_train_snap = MetricsManger.make_metrics_rsc_mul_symbol_grid(buy_factors, n_folds=6, \n",
    "    score_type=METRICSTYPE.SYSMBOL_R_SCORES_GOLDEN.value, ret_cnt_need=0, train_test_split=True, \n",
    "    use_last_test=False, force_one_process=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 使用hdf5将这些返回数据缓存，可以只从缓存加载，节约时间\n",
    "orders_pd_train_snap = ZCommonUtil.load_hdf5('orders_pd_train_snap', 'orders_pd_train_snap')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>buy Date</th>\n",
       "      <th>buy Price</th>\n",
       "      <th>buy Cnt</th>\n",
       "      <th>buyFactor</th>\n",
       "      <th>Sell Date</th>\n",
       "      <th>Sell Price</th>\n",
       "      <th>Sell Type</th>\n",
       "      <th>Symbol</th>\n",
       "      <th>MaxLoss</th>\n",
       "      <th>ExtraInfo</th>\n",
       "      <th>key</th>\n",
       "      <th>profit</th>\n",
       "      <th>result</th>\n",
       "      <th>sell_type_extra</th>\n",
       "      <th>R</th>\n",
       "      <th>sharpe</th>\n",
       "      <th>profit_cg</th>\n",
       "      <th>profit_cg_hunder</th>\n",
       "      <th>keep_days</th>\n",
       "      <th>lowBkCnt</th>\n",
       "      <th>his_miss_days</th>\n",
       "      <th>atr_std</th>\n",
       "      <th>deg_hisWindowPd</th>\n",
       "      <th>deg_windowPd</th>\n",
       "      <th>deg_60WindowPd</th>\n",
       "      <th>jump_power</th>\n",
       "      <th>diff_days</th>\n",
       "      <th>wave_score1</th>\n",
       "      <th>wave_score2</th>\n",
       "      <th>wave_score3</th>\n",
       "      <th>snap</th>\n",
       "      <th>symbol_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2012-08-28</th>\n",
       "      <td>20120828</td>\n",
       "      <td>26.79</td>\n",
       "      <td>7465</td>\n",
       "      <td>BuyGoldenFactorClass:42</td>\n",
       "      <td>20120905</td>\n",
       "      <td>24.460</td>\n",
       "      <td>loss</td>\n",
       "      <td>usMSB</td>\n",
       "      <td>19.17</td>\n",
       "      <td>{'wave_score3': -0.44112256310114328, 'wave_sc...</td>\n",
       "      <td>498</td>\n",
       "      <td>-17393.45</td>\n",
       "      <td>-1</td>\n",
       "      <td>BuyGoldenFactorClass:42: lossPrice :below200</td>\n",
       "      <td>-0.305774</td>\n",
       "      <td>-4.42247</td>\n",
       "      <td>-0.086973</td>\n",
       "      <td>-8.697275</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.985224</td>\n",
       "      <td>3.939047</td>\n",
       "      <td>-1.154262</td>\n",
       "      <td>3.300417</td>\n",
       "      <td>-1.111737</td>\n",
       "      <td>35</td>\n",
       "      <td>-0.121029</td>\n",
       "      <td>-0.200734</td>\n",
       "      <td>-0.441123</td>\n",
       "      <td>/Users/Bailey/Desktop/my_work/abu/data/save_pn...</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-08-30</th>\n",
       "      <td>20120830</td>\n",
       "      <td>21.74</td>\n",
       "      <td>9199</td>\n",
       "      <td>BuyGoldenFactorClass:42</td>\n",
       "      <td>20120905</td>\n",
       "      <td>21.365</td>\n",
       "      <td>loss</td>\n",
       "      <td>usINDY</td>\n",
       "      <td>19.16</td>\n",
       "      <td>{'wave_score3': -0.89585840169126973, 'wave_sc...</td>\n",
       "      <td>500</td>\n",
       "      <td>-3449.62</td>\n",
       "      <td>-1</td>\n",
       "      <td>mv_pre_atr preClose 2.0</td>\n",
       "      <td>-0.145349</td>\n",
       "      <td>0.329131</td>\n",
       "      <td>-0.017249</td>\n",
       "      <td>-1.724929</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.089305</td>\n",
       "      <td>-5.110612</td>\n",
       "      <td>1.030533</td>\n",
       "      <td>3.618710</td>\n",
       "      <td>-4.396298</td>\n",
       "      <td>37</td>\n",
       "      <td>-0.552357</td>\n",
       "      <td>-0.595897</td>\n",
       "      <td>-0.895858</td>\n",
       "      <td>/Users/Bailey/Desktop/my_work/abu/data/save_pn...</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            buy Date  buy Price  buy Cnt                buyFactor Sell Date  \\\n",
       "2012-08-28  20120828      26.79     7465  BuyGoldenFactorClass:42  20120905   \n",
       "2012-08-30  20120830      21.74     9199  BuyGoldenFactorClass:42  20120905   \n",
       "\n",
       "            Sell Price Sell Type  Symbol  MaxLoss  \\\n",
       "2012-08-28      24.460      loss   usMSB    19.17   \n",
       "2012-08-30      21.365      loss  usINDY    19.16   \n",
       "\n",
       "                                                    ExtraInfo  key    profit  \\\n",
       "2012-08-28  {'wave_score3': -0.44112256310114328, 'wave_sc...  498 -17393.45   \n",
       "2012-08-30  {'wave_score3': -0.89585840169126973, 'wave_sc...  500  -3449.62   \n",
       "\n",
       "            result                               sell_type_extra         R  \\\n",
       "2012-08-28      -1  BuyGoldenFactorClass:42: lossPrice :below200 -0.305774   \n",
       "2012-08-30      -1                       mv_pre_atr preClose 2.0 -0.145349   \n",
       "\n",
       "              sharpe  profit_cg  profit_cg_hunder  keep_days  lowBkCnt  \\\n",
       "2012-08-28  -4.42247  -0.086973         -8.697275          8         1   \n",
       "2012-08-30  0.329131  -0.017249         -1.724929          6         1   \n",
       "\n",
       "            his_miss_days   atr_std  deg_hisWindowPd  deg_windowPd  \\\n",
       "2012-08-28              0  0.985224         3.939047     -1.154262   \n",
       "2012-08-30              0  0.089305        -5.110612      1.030533   \n",
       "\n",
       "            deg_60WindowPd  jump_power  diff_days  wave_score1  wave_score2  \\\n",
       "2012-08-28        3.300417   -1.111737         35    -0.121029    -0.200734   \n",
       "2012-08-30        3.618710   -4.396298         37    -0.552357    -0.595897   \n",
       "\n",
       "            wave_score3                                               snap  \\\n",
       "2012-08-28    -0.441123  /Users/Bailey/Desktop/my_work/abu/data/save_pn...   \n",
       "2012-08-30    -0.895858  /Users/Bailey/Desktop/my_work/abu/data/save_pn...   \n",
       "\n",
       "            symbol_index  \n",
       "2012-08-28             6  \n",
       "2012-08-30             6  "
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.options.display.max_columns = 40\n",
    "orders_pd_train_snap.tail(2) # 看看添加了提供机器学习数据的pandas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "## 下面生成测试集数据\n",
    "\n",
    "        BuyGoldenFactor.g_enable_fiter = False\n",
    "        use_last_test=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "BuyGoldenFactor.g_enable_fiter = False\n",
    "BuyGoldenFactor.g_enable_snap = True\n",
    "buy_factors = [{'XD': 42, 'class': BuyGoldenFactorClass, 'draw': True}]\n",
    "out, orders_pd_test = MetricsManger.make_metrics_rsc_mul_symbol_grid(buy_factors, n_folds=6, \n",
    "    score_type=METRICSTYPE.SYSMBOL_R_SCORES_GOLDEN.value, ret_cnt_need=0, train_test_split=False, \n",
    "    use_last_test=True, force_one_process=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4956, 32)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orders_pd_test_snap = ZCommonUtil.load_hdf5('orders_pd_test_snap', 'orders_pd_test_snap')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((44412, 32), (4956, 32))"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orders_pd_train_snap.shape, orders_pd_test_snap.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 对训练集及测试集的结果进行大体的度量：主要就是看看**‘非均衡胜负收益’带来的必然’非均衡胜负比例‘，目标由’因子‘的能力解决一部分**效果怎么样"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from UmpMain import UmpMainClass\n",
    "from MlFiterGoldenPd import MlFiterGoldenPdClass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "all fit order = (41225, 32)\n",
      "win rate = 0.493559733172\n",
      "profit_cg.sum() = 281.687935222\n",
      "win mean = 0.0770327160087 loss_mean = -0.061738192442 \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAEJCAYAAABrHbdyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1315d30d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_ump = UmpMainClass(orders_pd_train_snap, MlFiterGoldenPdClass)\n",
    "train_ump.show_general()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "all fit order = (4612, 32)\n",
      "win rate = 0.5\n",
      "profit_cg.sum() = 37.4002795407\n",
      "win mean = 0.076550446692 loss_mean = -0.0605681543452 \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAW8AAAEJCAYAAABbkaZTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XecVNX5+PHP9sIusMDSe9kjHYSgIE3EWCKx5JsYTYyK\nJRqNGkHFGH9RExMbaiyxY02M0WgUYxeVoqgUpT/03nZhF9heZn5/3DuzM7uzbXbKnd3n/Xr5cm5/\n5uzwzJlzzz0nzu12o5RSKrbERzsApZRSTafJWymlYpAmb6WUikGavJVSKgZp8lZKqRikyVsppWJQ\nYrQDUMqXMeZi4C5gHfA1sFlEXjHG3A58JyLzoxqgUg6hyVs5za+AW0XknzXWTwPWRiEepRwpTh/S\nUaFgjJkC3A/sAfoDxcAlIiLGmOeBDvb6d4G/Ao8DowAX8D5wm338FcBB4BF7+xqgBLjXXn+jiLxd\nTxwnAH8D0oFy4CYR+cwYcyZwD1AJfA9MB04SkZ0NHD9bRD43xriATiJy2N7PBXQChtvvZy8w1H7f\nfwSuA3KAN0XkRmNMG+B5YKD9npeLyK/tcntMRIb7lONjIjLcGPNHYID9XzfgG+Aj4GKgL3CziLzW\nwJ9GtVDa5q1CaTRwv4iMBF4AXvHZliYiw0XkVqzEnGcnrLFYSXqWiNwILMNKmH+zj3OLyN/t9Tc1\nkLgTgbeAO0RkBHAl8LAxpgPwEnChiBwPfAZ0b+TxfzPGxAW4nG+tZyxwl4gMBg4Ac4AzgDHANcaY\nrsC5QIZ9/XH29foHOFfN5ZOA04AhwKnAYBGZAvwWq3lJtVKavFUofS8iX9qv5wGjjDFZ9vJin/3O\nAB4DEJEK4El7nUegZNkYw4FKEfnAPvcK+4tkMrBWRNbY618CjjX2eBEJ9PPUN8ZtIrLKfr0F+ExE\nqkTkEHAU61fHYmCoMeYzrOT+sIhsbcR7+kRECkWkFKt2/4HPdbLqPky1dJq8VShV+ryOx0pwVfZy\noc+2msk5HkgK0fX9Eq0xZqi9vuZn3dXY440xCfZinL0uqcZ+ZTXOU1HzxCKyHavJ5C9AJvCpMeY8\n+zy+5ZFc49AGz61aJ03eKpRGG2OG2a+vBJaIyNEA+30IXANgjEmx9/2ogXNX0nCCF8BtjDnFPvfx\nwKfAUmCQJzZjzE+AdtRurqjr+His9vax9n4/aSCOWowxVwEviMjHdtPRh8AwIBfobYzpZDfPnNOE\n0wb7C0W1AJq8VSjtB+42xqwCfgxcZK+vmSSvB7oYY1Zj3Txcj1UjDbSvx3zgAWPMRXVsR0TKgfOA\nO4wxK4G/A+eKSB5wIfCyMWYZ8EOsL4PiRh5fYcf8d/v4kcC+ekuimuf9vAQkGGPWGWO+xap9/01E\n1gNPAcuBL7GaRho6V13LqhXR3iYqJOxeEo/aN/ocxRiTCfwB+KOIlBpjRgPvikiPKIemVNAa3c/b\nGNMZ647/dKx2zBew2g3XiMg1YYlOqRqMMbOBX+Bf64yzl+8XkVdrHiMix4wx5cAyY0wFVhfAn0Yi\nXqXCpVE1b7sL1b+xuiv9GKs/7gMissgY8wTwQX1duJRSSoVWY9u8HwCewGqPiwOOF5FF9rb3sWrj\nSimlIqTB5G2MuQQ4KCIfU3132/e4Y1h37pVSSkVIY9q8LwVcxphTse6yvwRk+2zPBAoaOkllZZU7\nMTGhod2UUkr5C9gltMHkbT+KC4AxZgFwFXC/MWayiCzEejJuQUPnyc8vbmiXsMrOziQ3N9BDda0v\nDifE4JQ4nBCDU+JwQgwaR+A4Agl2VMHZwDP2k2brgTeCPI9SSqkgNCl5i8g0n8WpoQ1FKaVUY+kT\nlkopFYM0eSulVAzS5K2UUjFIk7dSSsUgTd5KKRWDNHkrpZQDlFVUsXnPEVxuNzv2H2PfoaJ699fZ\n45VSKsr+vWAzH3yzs9b6C6YP4sIzhgQ8RmveSikVYS63G5fPiK6BEjfAq59sqvMcWvNWSqkI2p1b\nyF9fWUH7jGTuumwcnyzb7d3Wtk0yR4vK6dU5g0E929G7S+BH40GTt1JKRdSarYcpKaukpKySK+77\n3G/bzDOPY8SATo06jzabKKVUBG3fH2hObkuHtqmNPo/WvJVSKgLcbjdvLtzKN+sP1trWq3MGY3Ky\n6Zmd0ejzafJWqpUqKq3g02W7+e/ibTz824lkZzd8jApebkEJ//tqR631j94wiTapSU0+nyZvpVqp\n595dz3eb8wC468Vveeh3U6MbUAtX5bJ6lwzr14HLzxrC91vyGD+0K4kJwbVeN5i8jTHxwDOAwZot\n/iogGXgX2Gjv9oSIvB5UBEqpiKuodHkTN8Dho2V8s3Y/YwZ2jGJUrUOndqm0bZPMpBHdm3WextS8\nZwBuEZlojJkC/AWYD8wVkYeadXWlVMSVVVRx9dwvAOjdJYPTxvXmmfnrKC2vinJkqikarK+LyNvA\nlfZiXyAfGAOcZYz5whjzrDGmTfhCVEqF0pLV+7yvz53Un+4drX++67cfilZIKgiNamwREZcx5gXg\nb8A/gK+B2fb8lluBO8IVoFIqtPp2bQvAuMGdGTmwE26sttgvV+2r7zDlMI2+YSkilxhjOgPfAONF\nxPOXfgt4pKHjs7LSifbs8XVN5BlpTojDCTGAM+JwQgwQuTj25pcCcPhYGdnZmfZ1l5GWktDkGGbM\nehuA+XPPDmmMLfFvUuqy/p+alhyS8zbmhuUvgZ4icg9QinXT8k1jzHUi8i1wCrC8ofPo7PHOicMJ\nMTglDifEEOk4Nu6wmkc27z7ivWZqcgId2qY2KYbi0grv6337jwTda6KmlvY3yS0o4e6Xl3PdT0YA\nUFpS3qTzNmf2+DeB540xX9j7Xw/sAh4zxpQD+6luE1dKOcC1Dy2kuKySeXOm1dr21Zr9AFx65nHe\ndV2y0tl3uGkVLN+HTa68/3POmtCHGRP68f3mPApLKpg8sjvx8XFBvoOW45YnvwLgzy8tAyA5KTQt\nEA0mbxEpBs4PsGliSCJQSoWU2+2muKwSgJn3LPDb9v8uGcum3UcAaJue7LetvKKKo8Xlfuvzj5Wx\nakseb3y+hdt+NZauHdK92176UPyOf/fLHbz75Q6/7U/NnkJSlJtLo6G8ooqr7B49NXXOSgvJNXRs\nE6VamBI7cQdy1wtW7S8xIY7h/av7dO84YP2Mv+GRxcy8ZwG5BSUsWb2PWY8v4cUPhKLSSn7/9FK2\n7bPG5aiodHmPHdI3q87rHThc0qz3EotcLnediRsgOURfZpq8lWphPLXuLvXU8Cqr3PU2adzy5Fc8\n97/1tdb/6cVl5BaU8OsHPveum3X+qFr7TR/bE4Dt+6Pfdh1JLreb/321vd59xh4XmnEINHkr1dLY\nY/x3aJvKYzdMCrjLWRP6+i3Pveakek+ZnGilihkT+nrbcAFGDexEXFwcT82eAsDAHu14+qap9LO7\nI857bz35x8p4Zv46Zt6zgC17jgTzjmLGwu/38taibYB1E9hTbr5Sk0MzKomObaJUiB0pLOMfH29k\n+the5PRqH/Hrr952GID1O/JJT03yu2npdruJi6td487KTOGeayYy5/HFfuu7dEjnr1eeyM4Dx7jj\n+W+Z/+V2v+3DB1hNL0mJCX7XGXtcZ95evI2DBSXMenyJd/3dLy/nlON7ctaEPrTLSGn2e3WabXut\nZqWJw7sx80eDWbPtEE+/s45bLhxNZZWbDm1D9541eSsVYv9asJllkssyyeWqs4cybnCXiF3b5Xbz\n/lLrpuHvfjay1vZAidtjaP+O3gR86IjVF7xjO2t86e6dAj9EffLoHgHXJyXGc+mZx3HvP1fW2vbp\nit18umI3melJPPCbk0gKUDt1un9+stFvBpxnbp5KQnw8i+wHnaaMssYtGdavI49cH/jXT3PFXqkp\n5WCFJRV8ve6Ad/nJt9dG9PoVFS7yjpSS06u93w3JpurYLtWbuIGAfbifu+Xkes/Rs3Ptsal/+INe\n3tfHiit4bUHdczQ61Y79x/wSN8AV933OR9/sJCPNGtq1ri+7UNLkrVQIrbWbLFKTq3sUPP7mar/x\nRCLB9/qhcuP5/jX5+mrxAOkptX/Ynz9toN/yGru8YsnOg4Fvwv5rwWYKSyoCbgsHTd5KhdD7X1tN\nFr863dAm1UpeyzfmBuy5EWuG9auuyXdu33Bf5bi4OO6+4oRa63wdzC9psHeG06T53HA8Z2I/unVM\nr7VPShi+PGvSNm+lQuS5d9ex80AhAAO6tyMrM4Wi0rr7XIdDSbl1vVVbwjNC4Lw50ygsqSChkU9O\nduvYptZTnhecMohXP61uLvnPF1v50fi+oQwzYn48sR8nDuvKHJ8eONeeN5z4Bn6VhILWvJUKAZfb\nzRL7sfMzT+xDdvs0ducW+e2zeXf4u8nVbIsNh4y0JNICNIk01qk/6MW8OdP465UnetflHysLRWgR\n4XJbfTF/cWoO4P8rZOqo7hyfE5n55DR5KxUCl9/7mff1ScO7AnD7xWP99nngNavnRWl5JRt25PPl\nmn1s23cUlz09ltvtxm0nhmANredpR6fp0iGdk4ZZZVXfU6FOUuVyeW9CB/rCycqMXPdHbTZRrZbb\n7WbGrLdJiI/j6ZumNngDLhCXy83l91Un7jEmm2725Ab9urX127e8wsXStft5/v0Nfo+XJyXGk5GW\n5E0GgQaTaqy8o1YXv/Mm9w/6HJHU1W4v3p1bGJEeGs1V5jPbkOdLGuCUMT35dPlupowK3HUyHLTm\nrVotT/t0lcvt1wbbFBt25vstX3PucL/l5245mZlnDvYuPz1/nTdxZ2WmMGlEN7p3atOsZghflfa5\nsxtxQ9EJOrS1uiN+tymvgT2dZfSgTt4vabCaUObNmUbbNsn1HBVaWvNWrdaabdU39RZ9v48Lp+c0\n+tiKShevf7aZT5Zbbcx9umYy++e1x/iIi4tj4ohurN1+2K//90O/nUg7n3/o2dmZXPTH90M2XGgE\n7peFhGeEvaXrDnDFjCFB/fpprTR5q1bF7Xbz4Te7+Pdnm/3Wl1VUcc8/VjDnF8d71+3JLSQ5KSFg\nLfbdL7d7EzfAaeN60SY1qc7rXjljiF/ybldHDa04wr1Tom1A93a0z0imoLCcW59aSnx8HJUuN6Vl\nlfTo1IabLhwdkZ4bsagxM+nEA88ABmsWnauAMuAFe3mNiFwTxhiVajLP48t3zhxHL/tJv90HC7n7\n5eWUVVS3Ww4b0JE1dre6jbsK2LL3CD06tWHlpjyemb8OgEvPOI5JI7v7nd8zxkdCfByzfz4K07v+\nG4VxcXH0zG5TqweKr45tU9my9yhlFVWkBFkDl10FQR0XTTeeP4q3F21j3Y58khLiSE1NoryyCtlV\nQGWlK2S/RlqaxtS8ZwBuEZlojJkC/AWIA34vIouMMU8YY862Z5lXKurKKqq8Xeb+OO8b7w3A595b\n75e475w5jmE5ndmz7wjX/W0RAHe/VHtGv+ff38D8L7fzp8tPICUpgc0+I+M9ddPURtcM77rshHq3\np9dTc/fw9EYpr3SRkpTgnWwhIy2J6386wtt27GlLjgU9szO45rzqewXZ2Znc+tgi1mw7zIadBYwY\nEPxj/i1ZY2bSedsYM99e7APkA9NFZJG97n3gVECTt4q64tIKrn14kd86z0h65Xbinji8GxefYUiI\njyc5KYGMtCTuu2o8N/s8aFFT3pFSrp77BdeeN5xvN1jTf007vkdIf9J7ZnHfk1tE/+5ta23/et0B\nnnon8FgphSUV3i+entkZDOzRLmRxRcO67daN4Idf/75W75u8ghLSUhNJS06M+DRrnh5BThhMq1Ft\n3iLiMsa8AJwD/BQrWXscA2L7k6JiWnFpJdc+vLDO7a9+usl7MzIzPYmZPxpca59O7dMY3CeL9Tv8\ne49cMWMIH32zyzvTzGNvrvZua8oNzsaoqrKS9+JVe2sl75Ubc+tM3DX16+aMmdeb4+Tje/Dp8toP\nHN342GIKCsutfUb34KLTTETj2rbP+hz06RL9Mm70DUsRucQY0xn4FvC9g5MJNNjQlpWVTmKU57Kr\naxbmSHNCHE6IAUITx9qttR8F/90Fx5OSlMA9L33LJ8t2s/C7vWRnpRMfH1frmp7l+66bzIxZ1T8g\n5889G4AfTx1EwbEyLrrjA7/junSpXTtujl+fN4IbHvqCCnftclnx/gYAfnrKIM4/1XDF3R97a4Fz\nr5/MH5780vugS1F5VdDl6pTPxXnTBnmTt29MnsQN8NnKPdz4y7G1jg2lmuWRkGz96hrQp0PUy6ox\nNyx/CfQUkXuAUqAKWGaMmSIiXwBnAAvqOwdAfn7TZqYOtezsTHJzoz8lkxPicEIMoYzjk6Xb/ZZ/\nf9EYb7PBzDMHM++99ZRXutiTW0indql+16wrhmdvObnW+tsvHsufXlzGbb8aw4Du7UJahtnZmWQm\nxxMXB/vzCv3OfbSonCXf7wVg+ujuHC0o5vr/G8Edz38LwLGjJTx6/STvw0Ibd+QHFZuTPhclRdVP\nL9YX04xZbzfroaaG4qh57cJCK65jR0siVlZ1fUk0pub9JvC8MeYLe//rgA3As8aYJGA98EaI4lSq\nyT6p8fPat7134ohuvPKxUF5hPbySZ08yUJf6EkG/bm3DlijA6pGSkZZEQY3Hrj9bucf72tPG69t9\nsX1mCvHxcZx+Qm8++Hont100JmwxRkqgG655R6I/mfFWe6YcJ3RfbMwNy2Lg/ACbpoY8GqWaoeaD\nLx5nntCH/y7eFoWImq5ndgbrd+Szbd9R7+P1a7dbY17feP5IEuKtG2VpKYnMmzMNl6t6IuGfnTyQ\n8yb3DzhxQizbuKuAgT3bMc8eVrddm2Qe+u1Eb0+bPXlFdO+YHvYHfNxut7ev/pHi8gb2Dr+W9VdW\nrVpdD76sCtAm7lSeG6Z/enEZZeVVbN17lM27j5CWkuA3nrZHzd4WLS1xAyxYsZtPvt3Fhp0FjB7U\niQev9Z8s+fZnv6710FU4eG5aAwzqGfm5SWtqeX9p1aoUlVozl9TXw6JLVu3B8p3KdxyUqx/8gj+/\ntAyAkrKqug5psS4+3epJkpyYwFuLtpGRlsTFpx/nrWGfOKR6btAPv9nFolV7wxrPXS9Yf4tfnW7o\n4YBBtDR5q5hWWGwl7zZpdT/gctFp1V36/uIzhrQTTRzRLeD6C6YPinAk0Td+aFeSE+NZvHofZRVV\nDO/f0W/gpyt/PNTvHsTz721o9pC6ddm276j3dUeHPAClyVvFrLLyKm59eikAa7bWPRdiarLVPjxv\nzjS6dnB+LfyME3vXWnfq2F4B9mzZkpMSKPcZOvertfsbPGZvXt3DDzTHq59UjzqZ44AmE9DkrWLY\n1Q9+Ee0QwqIx80O2RmdN6Btw/f1XT/C+DvX9jcNHS9mx/xjFZZWkpSQwb860iMxP2RiavFXM+e+i\nrd6eBh7h7MIXaROG+TedtKT31lS+w+yePi7wr4+O7VK99zw27AjdwFwH84uZ/fcvufOFb9mbV+S4\n+w6avFXMeWfJdr9lz42tliIpMZ6bLxgNaC18SN8O3iav+gbuuuVCayjf1VsPhWxY3c0OH6FRk7eK\nKTVvSJ01oQ+TawzX2hIc1yeLu684gb/+2tk3WJ3Cd9jYax9e6B08rDlWb/af3cd32jMn0OStYkZh\nSQWX2RP9DuhhPe143uQBLXb2lW4d27TY9xYOP506wPv6if+uada53G437y6xHuxKSUrglDE9+eWp\nzvqFp8lbxQzPmNsA+w9Fd6wc5TxnnNjH7/7AgWaMpzT77196Xz9+42R+cWqOY25UemjyVjHDt4/v\nHy/9QRQjUU72s5MHAtUTTAfDM2LjmJxsR4xjEogmbxUTlqzex9EiazyJx26YTKd2rftGnqpbSpKV\n1tY0o9tgp3bWgzi+M/w4jSZv5XiVVS6eswcliouD9FSdN1vVzTPi4qJV+xr1YE8ghSUV9Aswm5GT\n6L8C5VgHDhd7n6D0uPZc59aElDN08uleuXjVPsYPbVovkbLyKkrLq8jKdMZj8HVpzGQMicA8oC+Q\nDNwN7ALeBTbauz0hIq+HKUbVCtV8CMdjdE52hCNRsaZrh3QGdG/Llr1H2VBjWruG+FYY2memhCO8\nkGlMs8kvgTwRmYw1a85jwPHAXBGZZv+niVuFTF2DC00f2zPCkahYdeb4PgC4sSalbqw3Pt/ifd0+\nI/aT97+B2332rwDGAGcZY74wxjxrjIn++Iiqxfh4WfXMOHN+cbz3dagn/FUtVxufpzGvfXhRPXtW\nKymrZPnGXO/ymMGdQx5XKDWYvEWkWESKjDGZwOvAH4BvgNkiMgXYCtwR1ihVq7J17xEAbrlwNDm9\n2jNhWFdOOV5r3arxBvZs57dcUFhWx56WLXuOcM1DC/3WjRjo7Ca6Rt2wNMb0wprL8jER+Zcxpp2I\nHLE3vwU80tA5dPb4ak6IwwkxQOA4Su35JseP6klCQjy3XnpCxGOIBifE4YQYIDRxTBzZncX2xM0f\nL9/Db/5vZJ37flhjHtS+9hR0TimPQBpzw7IL8CFwjYh8Zq/+0BhzrYgsA04Bljd0Hp093jlxOCGG\n+uIoKrH6c+flFdaa5itSMUSaE+JwQgyhjMOTuAHyj9Q/2/ue/da22y8ey75DRYwbbM3S45TyCKQx\nNe9bgfbA7caY/4d1D+B3wMPGmHJgP3BliOJUitz8EjLTk3Dog20qRjwxawpXz7XGfD9SVO43WXNN\nq7bkkZ6SSPeObbwTPztdY2aPvwG4IcCmiaEPR7U0ry3YxOGjZVx9zrBG7b83r4ij9tRmOiiTao6U\npASeuXkqtz39Net35HPTE19yx6U/IDO9epiFg/nFrNiY5/3MOW38kvroE5YqrD78ZhffbjhIaXnd\nYyyXllfyzuJtLJdc/vDs1xGMTrV0CfHx3Hj+SDLSksg/VsYNjy5md6415snjb61mzlNLIzLzfDho\n8lYR8cL7G+rc9psHF/Lfxdt4/K3V3nWewYWUaq7OWencfcUJDOzRDrcbPlu5h715RSyXXL/9fLul\nxgJ9PF6FTZHPwxHfrD9ITq/dTGugy985k/oxfUwvHb9EhVRmejIXn264/blv+GzFHj5bsce77bzJ\n/enQNoWcXs6YWLix9F+ICqmyiioO5pfQq3MGf37JvxPSKx9t5MQhXetMzK15rkYVfj2yM2qtG5OT\nzY/G94nJ+yuavFXIbNpdwF9fWVFr/YwJfZn/5XYAXvxgg9/NS1cdj8IrFQ5dstI4kF/iXb7yx0Nj\nMnGDtnmrENm4K3DiBjh3cn8eu2EyAN9uOMh7S3ewJ7eQopIK/vTCskiGqVq5v/56PO0yqnubJCXG\nbgrUmrdqtopKF/f+I3Di9vBtKnnj8y1+AwApFUl3zRzH9Y8sjnYYzabJWzXb7txCPI0fT980lcQE\nqzaTd6SEJJ8hEU4c2oWlaw94l0fnZLPSHghI27tVpGSmJ7eIz5smb9VsW/ceBaB7pzbexA3Umqrs\nyhlDycpM4cQhXenVOYPs7EzmPLaI1Bh6MEIpp9DkrZrN5bLq3edO6t/gvj+d6t9/e9b5o8ISk1It\nXey21iulVCumyVsppWKQJm/VbHVNW6aUCh9N3qrZducWAZCSrB8npSIl2Nnj1wEvAC5gjYhcE74Q\nlZMdLS5n8ep9APQM8PixUio8mjp7/OlYs8c/CPzensMy3hhzdhhjVFHy0gcbmHnPAh587Ts27MgP\nuM+xImvWmyF9sxw/27ZSLUlTZ49PACqB40XEMyXz+8D0MMSmouzz76xppNZsO8x9r65k5abcOvft\n0iE9UmEppQhu9vjbAN+RXI4B7QIerGKS2+1m5j0Laq3fsKMgCtEopQJp1B0me/b4BcCLIvIvrLZu\nj0xA/1W3IJv3HPFbHj+0KwAfL9vFzHsWUOVy4XK7cbvdFJdZM+SUltU9U45SKvSCnT1+pTFmsogs\nBM7ASuz1yspKJzExuo9B1zULc6Q5IY76YnjhA/G+nj/3bKpcbs656R3vukffXMOqzXl+xxQUVQT1\nvpxeFpHkhDicEANoHI0R7Ozx1wOPGmOSgPXAGw2dJD+/uDlxNlt2dia5uceiGoNT4mgohr151hx/\nz91ysne/nF7t2bjL+oFVM3EDnPaDnk1+X7FQFq0pDifEoHEEjiOQ5sweP7V5ISmn2rzbajbxHaT+\nlgtHU1pexQ2PLqai0lXrmOH9O0YsPqWUDkzVquUfK+P1zzaT06s9k0Z2IyG+7lsgcXFxpKUk0qFt\nKgcOV/+KeubmqfUep5QKD03erdCMWW/7LS9dd4CPl+3i7itOZOcB62diny6Bf6rd/qsxXPvwIu+y\nJm6lokOTtwJg3yGrNp1bUArA+KFdAu6XnprEc7eczOJV+zjR7oWilIo8rTa1QFUuF799eCG3PbPU\nb72rjv7bHt+sr57lpr5JWePi4pg0sntMz/+nVKzTmncL9Oh/VlNUWklRaSX7DhXRrWMbAP71ySa/\n/ebNmUZBYRlPv7OWDTsLePLttd5tMTqhtlKthladWpgNO/JZteWQd/m2Z772vu6e3cb7+vcXjQGg\nfUYKsy8YXes8Q/t1CGOUSqnm0uTdgrjdbl7+SGqt//CbnQCkp1g/tK46bwQDe1SPaBBfo5r92A2T\nvLV1pZQzafJuQfKPlbHvUDHD+nfgjkt/4F3/2oLNfLJsV73Heh6Bf/aWk0lPTQprnEqp5tM27xbg\nsxW7efmjjd7l7PZp9O6SyfnTBvLags0A/POTTVwxY0id57hixpB6tyulnEVr3i2Ab+IGOG+yNYv7\naeN6+61/Zv46APIKSiITmFIqbLTm3YLMOn9UrRuNT8yawksfCF+t3e9dN1hvRioV87Tm3QL06NSG\njLSkgD1EUpISuGLGEIbZ27pkpTFuiD5co1Ss0+Qd444UlrEnr6jB/UorqiIQjVIqUjR5x7g12w4D\n0LVj/dOQVVW5ATiQr+3dSrUEmrxjXGm5VaOeNrpHvfudPbEvABdOHxTukJRSEdDoG5bGmBOAe0Tk\nZGPMKOBdwNPN4QkReT0cAar65do9Rzpn1V/zHjGgEw//diIZ6dqHW6mWoFHJ2xhzE3ARUGivGgPM\nFZGHwhU3mQKjAAAZnElEQVRYa1JYUkFGWnBJdd32fABSkhueYq5tm+SgrqGUcp7G1rw3A+cCL9vL\nY4AcY8w5wCbgehFp+K6ZquXahxZSXFbJBdMHcerYXk069u3F29ida32fdmugzVsp1bI0qs1bRN4C\nfKcH/xq4SUSmAFuBO0IfWuvgmX391U82sWJjbqOPe3/pDt5evM27XHN8EqVUyxbsQzr/FZEj9uu3\ngEcaOkBnj6/mG8ep43rzsT1w1GNvrmb+3LMbPL64tILXP9/iXf7vfTNISGjavWcnlkVrjgGcEYcT\nYgCNozGCTd4fGmOuFZFlwCnA8oYO0NnjA8dRVenf/3rj1jyyMlPqPP7w0VJm//1L73JGWhKHDzet\nxcqpZdFaY3BKHE6IQeMIHEcgwSbvq4FHjTHlwH7gyiDPo2xtUhMpKq1k1uNLuP/qCXRsl1prn0Cz\n4Nx71fhIhKeUcphGJ28R2QFMsF+vBCaGK6jWqHeXTNbvsHqO3PTEl/z4pL6cM6m/d/v2/UcDHpeW\nosPTKNUa6b98hzhnUj9v8gZ4Z8l2b/J+a+FW5n+53W//5245ud55JpVSLZsm7yhYsGI3bdOTGXtc\nZ++6tJRELj9rMM++ux6Ajm2tdu9jxeXexD19bE9+fFK/oPuEK6VaDk3eEbZh+2FescffvvyswX7b\nJgzrRofMVO57dSWHjpZx2zNLvWNzjxrYiQun50Q8XqWUM+nYJhF206OLvK89tWxfbrfb+3rfoWIe\nf2sNAAN6tA1/cEqpmKE17zApKaskPj6OB1/7jsSEeHp1ziAlqXY/d5fL7bfcIUAvE4D4eG3fVkpV\n0+QdBoG69PnejPT1+Xd7/Za7ZKXzyPWTyEhL8jtP23Qdl0QpVU2TdwT88oc5dGqXSkpSAvf+cyUA\nJx/fg89W7PHu41sDD3RDcuTATuEPVCkVM1pV8n71k018vGwXADeeP5Jh/TqG/Bqbdhd4Xz983URK\nyirp4jNc67//8iOOHinm8TdX+x2371Axvbv4P0k1b840Vm89RHpKovYwUUr5aTU3LDfuzPcmboAH\nX/s+LNd576sdAFxyxnG0TU/2S9xgdQmMj4vjjBP7+K3v1y3wI7DD+3dkQI92YYlVKRW7Wk3Ne9bf\nFkbkOrlHSklOimfi8G717jewRzueu+Vk3G5wud0kNnFgKaVU69YqMsbm3Uf8lpMSrbe9Y39oB51x\nu93szSuivMLVqN4hcXFxxMfHaeJWSjVZq8gasqu6p8e8OdPomd0GgC++21PXIUFZteVQSM+nlFJ1\nafHJ+0hROf/5YqvfupsvOJ6UpARWbz0ckmuUllfy+6eX8rc3VoXkfEop1ZAW1eYtO/O5958rmTSi\nG4tW7au1/UfjrZuEKckJ9OmaycZdBRQUltE+o+7xs+tTWeVixcZc5v1vPeWVrmbFrpRSTRHs7PED\ngBcAF7BGRK4JU3xN4ulDHShxg//wqYN6tmPjrgI++HonP5s2EID7/rmSC04ZRKf2qbRJrbtrXmWV\niyvv/7zO7f2766PsSqnwCnb2+AeB34vIImPME8aYs0Xk7XAF2Ri7DxbWu71dRjJTRnX3Lp86thf/\n+2oHH327i4++3UW7jGSOFJZz5wvfeveZN2darfNs3n2Ev7xS98RBF04fxJRRPYJ4B0op1XiNbfP2\nzB7vMUZEPCMsvQ9MD2lUQXjuvepBnu6/egK/OWcYndunAfDkrCm8cucZfrXptm38Hzc/Ulhe65wL\nv99ba13NxH3XZeO8rx/+7USmj+3l7c2ilFLh0qiat4i8ZYzxfarEtx/cMSCqT5EsXbvf2+2vXUYy\nHdul0rFdqt942cFYLrlMHmnV1iurXNzzjxV+25+95WTi4+KYe81JbN9/tNYXglJKhUuwNyx9785l\nAgV17egRztnjn55vDeA0uG8H7r12Yp0zzNScyLNT+zTyCkq8y5NG9eDmi8aybP0B7nx2KfEJcWRn\nZ+JyuTn7pnf8jvWd5T07O5Oc/o0fe8QJM1I7IQZwRhxOiAGcEYcTYgCNozGCTd4rjDGTRWQhcAZQ\nexi9GsI1e3xpeSUAQ/pmMfvno8jLC9z2HWgm6PuuGo/b7WbjrgI6tk2lU/s0cnOPkZVuFcv3m/KY\nMcu/KX/2z0cxpG+HoGeVdsKM1E6IwSlxOCEGp8ThhBg0jsBxBBJs8p4NPGOMSQLWA28EeZ5m22Q/\nPRlorOzGiIuLw/TO8lvXJjWRpMR4Kmp0/zt3cn+G9O0QXKBKKRVCwc4evwmYGqaYmuThf1sDTHXI\nDDyJQTAS4uN5avZUv/G0Lz3jOCaN7F7PUUopFTkx/5COZxTsC6YPCvm5n5o9hYpKF+n19PlWSqlo\niOnkfbSountfOKYJS0pMIClMN1mVUqo5YrpD8vb9RwE4fVzvKEeilFKRFdPJe+Mu62bl0P56E1Ep\n1brEdPL2NJukJmvThlKqdYnJNu/DR0v500vLOFJYTmJCfK2pxpRSqqWLyZr3uu353rFIrvvJcJ2c\nVynV6sRk8j50tBSAK2YMYVj/0M8Ar5RSTheTyfvtxdsA6JyVFuVIlFIqOmIyeXv0ys6IdghKKRUV\nMXfDcsf+YyTEx9EzO4PkIMczUUqpWBdzNe8Pv91JlcvNtDE6W41SqvWKqeTtdrvZtKuA1OQExg/t\nGu1wlFIqamIqee/OLeLQ0TJGDexEYkJMha6UUiEVUxlwyWprVvhenfVGpVKqdWvWDUtjzHLgiL24\nTUQua35IgVVWufj8uz1kpCUxdbS2dyulWregk7cxJgVARKaFLpy67TpYSHmFi5xe7UlLiblOMkop\nFVLNyYIjgTbGmA+BBOA2Efk6NGFVu/25r9mTW+RdXrP1cKgvoZRSMac5bd7FwP0ichpwNfAPY0xI\n29BXbsr1S9xKKaUszal5bwQ2gzWnpTHmENAN2BNo56ysdBKbOCtN2u6jtdYd1yerztmUGxLscaHm\nhDicEAM4Iw4nxADOiMMJMYDG0RjNSd4zgeHANcaY7kAmsK+unfPzi5t8garKSgDisOaqHDGgI9ec\nO4zc3GNNPld2dmZQx4WaE+JwQgxOicMJMTglDifEoHEEjiOQ5iTv54DnjTGLABcwU0RczThfnc6Z\n1I8ZJ/ULx6mVUiomBZ28RaQC+GUIY6ll9ZZDALTR8bqVUsqPox/SWbf9MMlJ8Uwc3i3aoSillKM4\nOnkDJCXE6+iBSilVg+OTt1JKqdo0eSulVAzS5K2UUjFIk7dSSsUgTd5KKRWDHDs8X2WVi906rolS\nSgXk2Jq3jh6olFJ1c2zy7tA2BYDJI7tHORKllHIexyXvsooq1m+vrnUnJzkuRKWUijrHtHkfOlLK\nJ8t38eE3uwC45IzjohyRUko5V9STd5XLxVsLt/He0h1+6w8EMYSsUkq1FlFvk9iy52itxA2Qf7QM\nALc70hEppZTzBVXzNsbEAX/HmseyFLhcRLYGc64d+63BzuPj4vj7jZN5+SNhyer93mFg03WyYaWU\nqiXYmvc5QIqITABuBR4MNoBXP90EwK/PHkpyUgLdO7YBqptNOrVPDfbUSinVYgWbvCcCHwDYM8aP\nbegAdwPtH6MHdQIgPdWqabtd2l6ilFJ1CTZ5twWO+CxXNjRz/GX3fkZeQYnful8/8Ln3dWKC/+Fr\nt+cDUF4RlpnVlFIqpgXboHwUa8Jhj/jGzF9585NfMX/u2d7likrrkOTEeO8km5mZ/s0kq7Ye5uen\nDw4yTH9OmQnaCXE4IQZwRhxOiAGcEYcTYgCNozGCTd5LgLOAN4wxJwKrG3NQSnKCdzZml08zSmJC\nvHd9WWmF3zHjh3YOyQzOTpoJOtpxOCEGp8ThhBicEocTYtA4AscRSLDJ+y3gVGPMEnv50sYcVFZe\n5X2980B1ofy/S3/gfX18TrbfMScO6RpkiEop1XIFlbxFxA1c3ZwLZ6Yle193bp/mfZ2SlMB9V43n\n5ie/4uYLRjfnEkop1WJFvBO12+0mLi6ORav21rlPp/ZpzJszLYJRKaVUbIl48i6rqOKjb3bxzpLt\nAJwwpEukQ1BKqZgX8eT9mwcX0jmrupkkLtIBKKVUCxCxsU1uPH+k9/XB/Or+3hOG6w1JpZRqqogl\n72H9OvoN85qRlsRzt5zMsH4dIxWCUkq1GBFtNpkwrCtdO6STlZlCtk8PE6WUUk0T0eSdmBBPTq/2\nkbykUkq1SFEfz1sppVTTafJWSqkYpMlbKaVikCZvpZSKQZq8lVIqBmnyVkqpGKTJWymlYlDQ/byN\nMbuBjfbiVyJyW2hCUkop1ZCgkrcxZgCwXETObnBnpZRSIRdszXsM0NMYswAoBm4UkY0NHKOUUipE\nGkzexpiZwO8AN9YIrm7gGuAvIvIfY8xJwCvAuHAGqpRSqlqDyVtE5gHzfNcZY9KASnv7EmNMt/CE\np5RSKpA4t88s7o1ljLkHOCQi9xtjRgJPiMiEkEenlFIqoGDbvO8BXjHG/AioAC4JWURKKaUaFFTN\nWymlVHTpQzpKKRWDNHkrpVQM0uStlFIxSJO3UkrFoBaVvI0xE4wxPezXcVGMo4cnjijGMNwYk2S/\njmZZdDDGpNuvo/J5M8ZMjcZ1azLGdDPGdLdfR/Nvcpkx5pfRur4dwzBjzLnRjMGO43RjzLBoxxGM\nFtHbxBjzA+DPQCHQG/idiCw2xsSJSMTeoDEmGXgQGAYcBh4Rkc8jdX07hjHAX4ESoAC4X0TWRKks\nHgEGAPnANSKSG4U4+gHrgGki8lWkrhsgjg7AfOAfwNMiUhmFGKZiPS29DXhQRHba6yP2N7Ef8Lsb\nGAU8KSL/jsR1A8QxGHgK2AwcwCqP3GjEEqyWUvO+DJgvIj8BngWuA4hkkrD9HIgXkanAZ8DlEb4+\nwEzgPXvQsB3AJIhKWZyHVRanAnlYzwZEI45hwH7gF/YXSsTZtew2WENLDABOjEYcwB+AL0TkBuBE\nY8w4iNzfxC6Hm4FEEZkGrDbGHBeJawdwJvCiiMzE+nx0iFIcQYu55G2MiTPGxNs1TIwxmcA+YIW9\nSxLwnWffCMTTxxiTbS/2AQ7arxcAbe34wnVtb1nYr1OwEvZXdqI6DUg1xoz17B+uWOzz9zHGdLIX\n+wJ77ddbgDJjzKBwJVCfsvC8V89nuwPWF1p34DzP9nCzy6IjeJNjN2A9sAsYbIwZaYxpE8bre8vD\npyweBi43xnwAHAc8ZYy5xrN/GGPpY4zJtsthKzDMGPMYcBfwjDHmamNM1zBe3y9n2Nx2HO8A7YF/\nGmOu8uwfrlhCKeaSt/0BmAS8bIzpKSLHgIdF5Et7lxHAap99w8Zu134AmGivmgv8yX49HlhrxxcW\nPmXxEtBTRMqAuSLyNXAyVjkcBr4wxqSFszx8ymKyveo+EbnDGDMUmG7H8TQwOhzX9ymLF40xvUTE\nZW/qjfXTeC3wDPCTCHyJ1SwLTxxPYSXwm7E+J0GPp98Q3/IAetqrVwBLsP42dwFXAVcYY5LD9dnw\nKQvPL8CXsd73NhH5KVZZDAW6hOP69jW9/06MMb3s1WX2Nf8jIncC1wO/NcYkRuHXYVBiKnnb36Bp\nwK+AbOAiY0y8iByxt6dj1XA+tG9eXhiuOOyXZwMnAGOMMTkiUoz1jQ7wQ+Bte//TjTFtQx2DT1l0\nxmoWSBCRKgAR+VBELheRF4H/YdW0Qi5AWRxvjDGe5Ckia0XkDBH5A1atK+T/SOsqC3tdb6wvt15Y\nTVnbwpioApWFp9yzgCeBOVi/DL8CUsMVR83ysDflAfdiJXDsL/mNWL8YQx6D/dK3LIba624A/mXH\n8BXQw44z5AKUxUX2pg+BFKCrHcdiYBNgwhFHOITtmz9U7J9TlwGfACIiBfZPnblYN8Q+AFbau/8Q\n6x/JnVgfmEdCHMt5WDWnvcAR+7/bgH7AeGPMZhFx2TfJXMBoY8xfgS/s+Jt7/YbK4kNgpbF6mZwH\nZGD9EknFSpwh00BZnGiXRZUx5mSsn6XlWIk0JDeFGlEWH4nICmPMduB/IvJfO3n80RjzmucLP0Sx\n1FcWJwAbgATgP1jJszdWAuuHdbMsFDHUWx7GmA9EZKUxZg9wqTEmg+oa+Y5QxGDHUV9Z/MAYs8H+\nu8ww1thIO7CaOneFMIbGlMUKY8y/gTONMQ9g/VtJBLaHKo5wc3RvE2PMFKwC/wjrV0Iq1t3ybBE5\naIz5I9a36U0iUmyM+Q3WB+VeEQlJ4rZrEF3sOLKwfnoOFZFzjTFdROSAMeZXwGDgLRH5xhgzHqt2\n8zzwkIisCUEcjS2L67E+hOOxalxrROTh5l7fjqEpZfEfEVlmjJkEXGrH+4CIrKjr/E2IozFl0R24\nylPLNiHuUdGEshiGdWNsbY3jO4lIXohiacpnIwmrCeE8YEMoPhtBfi6mY30+PZ+L5c2Nw46lMWXR\nDbja3t4POB84FqqcESmObDYxxnh+TrYHXhGR3wMPYd2xv0lEPDcF78e6e3+6vfw1MDKEidvT/tUF\nSBGRM+2f/92MMVeJiKfW9D5WzXKCsW5Q7gFOFZHLPN30gm1nDaIsfiwipSLyGVbyClXibmpZTDTG\ntBeRRcD1InKhXdsJur25iWXRB+snOxDa+x9NLIsSYLoxpr19bIIdT7MTd5CfjRIR+Qj4TYgSd7Cf\ni0+Aq0XkAhFZ3tz7EE0si77AOSJSJSKbsSaWianEDQ5L3sbqNfE6cLexOs53x/qJif0heBI41RjT\nxV5XDLyF9TM9XkSWh7A2cwMw1xgzGqs2sd2n/fJ3wE3GmEQ7jlysdsyeQEcR2Skin9rnSRARd1OT\nRzPK4gSfuELSlzjIsugBeHpbHLPPEx9MEm1GWYz3JMtQacbnooO9rioEMQRbHuN8YnMFPHnT4mju\n56LUPk9Qnwv72Ob8O/F8kTq3+aEejmk2McZMwOo69CDWT5lfADOwmh/OFJGtdjvdncAHIvJxmOLI\nxLpDvxOrBj0Aq9ZwJlab5UIRKTXGvAqsFJH77OMSgFQRKQpBDFoW1TFoWfjHEfXy0LJwhqjXvH1+\nLnUD9ojIeyLyOFAKFAH/BO4yVle3QqxazOYAx4fKYKCTiNwgIvfbcW3A6h0wBevJMOzldZ6D7J9g\nRc1sFtCysGlZ+HNYeWhZOEDUat41byAZYwYBcSKy0RgzCviDiPyfve3vQBUwEutDMhvrBkO4unxd\nhfWkZiJWN7M/YD18cwlW74GjWE/JXSQiq0JwPS2L6utpWfhf05HloWXhAG63O+L/5eTkxPu8Tgiw\n/d6cnJxb7deZOTk5bXJyckxOTs6EcMVhL8fVWJ6ek5Pzsc9yx5ycnJE5OTmX5uTkJGlZaFmEqyyc\nUh5aFs79L5o17zTgt1jtY0vtdQli9Q1+BHgF68m8acCvRGRv3Wdr8rXjoPpGhf3zqsR3u4i47dpF\nEbAGqwviIyKy0Gc/70MxzYxHy6L6PFoW/jFFpTy0LJwvIsnbvpvs8lkehDVQUZmIXFhj31SsD8Nq\n4E2s0b4KwxTXIOAvwCHgdbF7iNjbErEermmD9RTaM743PGr+hGvCNbUsqo/TsvC/ruPKQ8vCucL+\nhKXvN68x5gSsYVuPYT2BVRDgkD5YYyE8LyIbwhRHHNYIgL/G6guaDvzeGHNQRFbbH5xKY8xO4EsR\nebTm+YJMVloWgWNo1WURII6olIeWRWwJS83bGNMN+CVWZ/l99vJ9WJ3jS7C+ydthPen1hogsDeXP\nqwZi64P1KO61wGUiMtJePw/YLCJ/8dnX++1fsybQhOtpWVSfQ8vC/5qOLA8ti9gQrq6Cl2D9xPmJ\nvfwz4GMRmYQ1tsU5WGNt7AF+ZqyntEL6RzDVT1x5lqcZY97G6hf6O6w2uv8ZYzwD1azAGmnMS6xx\nSuLsn3/BPtRwCVoWHpegZeHrEqJYHloWsS1kNW+7MCvt18OwvjULsTrIdwTGAkOwuu1cijW2cD7W\nKHzviEhFiOLobV8zEXCJyMXGGkP6WTsmT6f9L4Dvsb7RlwODgGtF5PsQxKBlUR2DloV/HFEvDy2L\nlqHZbd7GGsHrHeAtY8xTInIYq+C3At8Cl4vI74z1NNS/gE+Bn2I9TvuGiOQ3NwafWM4AZmE9/voa\n1pjWAMOxOvRPwfpmfwfYjTX62cdAd/sbvrnX17Kovr6WhX8MjigPLYuWIxTNJpVYTzZdhfVzC2AR\n0B9rOMh4Y8xP7NdPY/0h5orIfWH4I4zFajN7XKwxToqNNb/lSiAZazCaaVjteT/C+jn2FpBpjBkR\ngutrWVTTsvDnlPLQsmghmt1sYt+V/iHWuAJdsMY4+B5oCxQDnYD/w2rTGiNhnJDXWJMvPIb1Bz8F\na7D1PKwxGB7FesrKDaRhjST2rbEmhj0NWCL2hKzNuL6WRfX1tSz8Y3BEeWhZtBwhafM21nCXl2MN\nDvMy1mD4W7BmMd+HNVjMq+LTyT9cjDVOcBugSkTetdctA07Fak8bLCF6XLeO62tZVF9fy8I/BkeU\nh5ZFyxDKG5bjsO5Q34o1e8ufsTr1/6neA0PMWJPw9hSRLfbyLKw58q4VazhIz35Bd2tqRAxaFtXn\n1rLwjyPq5aFl0TKEMnknA9cBQ0RkprEmB94dkpM3LY4OWHeru2GNJrYS+LOI7IlgDFoW1TFoWfjH\nEfXy0LJoGUL6kI6xHmEdA7wmURy9y/5JNhY4IiLf2uvCWosIEIOWRXUMWhb+cUS9PLQsYp9jJmMI\np2h8KJ1Ky6KalkU1LYvY0yqSt1JKtTRRn0lHKaVU02nyVkqpGKTJWymlYpAmb6WUikGavJVSKgZp\n8lZKqRikyVsppWLQ/wdJlkP5Vysu8QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x131d65850>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_ump = UmpMainClass(orders_pd_test_snap, MlFiterGoldenPdClass)\n",
    "test_ump.show_general()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "显示在所有有结果的单子中胜率可以接近一半，且每次赢的0.07比没次输的0.06多，\n",
    "但是贪婪的我仍然认为有优化空间，这个空间就是利用这个因子产生的输出数据\n",
    "作为特征识别基础，提炼特征，指导交易，\n",
    "\n",
    "**’模式识别‘提升关键的一部分**\n",
    "\n",
    "下面我们看看如何做到的呢"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "首先确认ml封装库是否能正常运行"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from MlFiter import MlFiterClass\n",
    "test_fiter = MlFiterClass.create_test_fiter()\n",
    "test_fiter_more = MlFiterClass.create_test_more_fiter()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**学习曲线**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY8AAAEZCAYAAABvpam5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrfz7fiM/vgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10bef0410>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(0.80412095046015697, 0.028989030232112856)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_fiter_more.estimator.logistic_regression()\n",
    "test_fiter_more.plot_learning_curve()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**特征权重**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "******************DecisionTreeClassifier******************\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature</th>\n",
       "      <th>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>x0</td>\n",
       "      <td>0.013333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>x1</td>\n",
       "      <td>0.013333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>x3</td>\n",
       "      <td>0.422611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>x2</td>\n",
       "      <td>0.550723</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  feature  importance\n",
       "0      x0    0.013333\n",
       "1      x1    0.013333\n",
       "3      x3    0.422611\n",
       "2      x2    0.550723"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_fiter.importances_coef_pd()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**树逻辑图**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAEBCAYAAADbxHY7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDCvClC6DwWBYkf8DHoxitnBbJ0AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12613ec10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_fiter.plot_graphviz_tree()\n",
    "path = ZEnv.shell_cmd_result('pwd') + '/graphviz.png'\n",
    "!open $path"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**混淆矩阵**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[[453  96]\n",
      " [100 242]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEUCAYAAAC8piQPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nwaFoZlZwKJqZFRyKZmaF/wFD81rVBcRy+QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x126607c10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_fiter_more.plot_confusion_matrices()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**roc**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "DecisionTreeClassifier :roc\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiQAAAEZCAYAAACn/z6eAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1266f4390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_fiter_more.plot_roc_estimator()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**特征选择**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "RFE selection\n",
      "             ranking support\n",
      "SibSp              1    True\n",
      "Parch              1    True\n",
      "Cabin_No           1    True\n",
      "Cabin_Yes          6   False\n",
      "Embarked_C         2   False\n",
      "Embarked_Q         5   False\n",
      "Embarked_S         3   False\n",
      "Sex_female         8   False\n",
      "Sex_male           1    True\n",
      "Pclass_1           4   False\n",
      "Pclass_2           7   False\n",
      "Pclass_3           1    True\n",
      "Age_scaled         1    True\n",
      "Fare_scaled        1    True\n",
      "\n",
      "RFECV selection\n",
      "             ranking support\n",
      "SibSp              1    True\n",
      "Parch              2   False\n",
      "Cabin_No           1    True\n",
      "Cabin_Yes          7   False\n",
      "Embarked_C         3   False\n",
      "Embarked_Q         8   False\n",
      "Embarked_S         4   False\n",
      "Sex_female         1    True\n",
      "Sex_male           9   False\n",
      "Pclass_1           5   False\n",
      "Pclass_2           6   False\n",
      "Pclass_3           1    True\n",
      "Age_scaled         1    True\n",
      "Fare_scaled        1    True\n"
     ]
    }
   ],
   "source": [
    "test_fiter_more.feature_selection()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**决策边界**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAD9CAYAAAC2l2x5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1260ed090>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAD9CAYAAAC2l2x5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKRe8mVlJueDNzErKBW9mVlL/B2AplqS8HR9CAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1266e94d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_fiter.plot_decision_function()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**train test score**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "accuracy mean: 0.776812507093\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 0.75555556,  0.72222222,  0.73033708,  0.75280899,  0.83146067,\n",
       "        0.78651685,  0.80898876,  0.71910112,  0.82022472,  0.84090909])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_fiter_more.cross_val_accuracy_score()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**切换使用其它学习方法**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "accuracy mean: 0.810433548973\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 0.78888889,  0.78888889,  0.78651685,  0.80898876,  0.84269663,\n",
       "        0.78651685,  0.82022472,  0.82022472,  0.79775281,  0.86363636])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_fiter_more.estimator.adaboost_classifier()\n",
    "test_fiter_more.cross_val_accuracy_score()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 可以证明封装库没有问题，可以放心使用，下一步上真正的股票数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from MlFiterDegPd import MlFiterDegPdClass"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "由252，60， 42天组成的走势数剧作为特质开始"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>result</th>\n",
       "      <th>deg_hisWindowPd</th>\n",
       "      <th>deg_windowPd</th>\n",
       "      <th>deg_60WindowPd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>0.139976</td>\n",
       "      <td>0.070454</td>\n",
       "      <td>0.068291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>-0.409467</td>\n",
       "      <td>-0.388299</td>\n",
       "      <td>0.750276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>11.841444</td>\n",
       "      <td>-0.846906</td>\n",
       "      <td>2.738821</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>-17.266100</td>\n",
       "      <td>-3.578966</td>\n",
       "      <td>-3.861270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>1</td>\n",
       "      <td>-9.146706</td>\n",
       "      <td>4.145140</td>\n",
       "      <td>2.854837</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            result  deg_hisWindowPd  deg_windowPd  deg_60WindowPd\n",
       "2015-09-11       0         0.139976      0.070454        0.068291\n",
       "2015-09-11       0        -0.409467     -0.388299        0.750276\n",
       "2015-09-11       0        11.841444     -0.846906        2.738821\n",
       "2015-09-11       0       -17.266100     -3.578966       -3.861270\n",
       "2015-09-11       1        -9.146706      4.145140        2.854837"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deg = MlFiterDegPdClass(orderPd=orders_pd_train_snap)\n",
    "deg.df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 上来就来个强模型，svm\n",
    "deg().estimator.svc()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "accuracy mean: 0.491279520472\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 0.48920689,  0.51758428,  0.49987873,  0.49551298,  0.48168809,\n",
       "        0.4887218 ,  0.47610963,  0.47355653,  0.4853191 ,  0.50521718])"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deg().cross_val_accuracy_score()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**0.49的准确率，相当于是乱猜** 是不是特征选择的不好呢，换一些特征试试"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from MlFiterMainPd import MlFiterMainPdClass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>result</th>\n",
       "      <th>atr_std</th>\n",
       "      <th>deg_60WindowPd</th>\n",
       "      <th>wave_score1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>0.176934</td>\n",
       "      <td>0.068291</td>\n",
       "      <td>0.338000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>0.580693</td>\n",
       "      <td>0.750276</td>\n",
       "      <td>0.769249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>1.438411</td>\n",
       "      <td>2.738821</td>\n",
       "      <td>0.242423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>2.061893</td>\n",
       "      <td>-3.861270</td>\n",
       "      <td>0.930900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>1</td>\n",
       "      <td>1.022388</td>\n",
       "      <td>2.854837</td>\n",
       "      <td>1.650291</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            result   atr_std  deg_60WindowPd  wave_score1\n",
       "2015-09-11       0  0.176934        0.068291     0.338000\n",
       "2015-09-11       0  0.580693        0.750276     0.769249\n",
       "2015-09-11       0  1.438411        2.738821     0.242423\n",
       "2015-09-11       0  2.061893       -3.861270     0.930900\n",
       "2015-09-11       1  1.022388        2.854837     1.650291"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from MlFiterMainPd import MlFiterMainPdClass\n",
    "main = MlFiterMainPdClass(orderPd=orders_pd_train_snap)\n",
    "main.df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "accuracy mean: 0.493608245706\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 0.5008489 ,  0.49890856,  0.49793839,  0.48678147,  0.50303177,\n",
       "        0.47125879,  0.49648314,  0.49369238,  0.49332686,  0.49381218])"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "main().cross_val_accuracy_score()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**0.49的准确率，还是乱猜** 是不是特征太少了，构建一个特征多的试试"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>result</th>\n",
       "      <th>deg_hisWindowPd</th>\n",
       "      <th>deg_windowPd</th>\n",
       "      <th>deg_60WindowPd</th>\n",
       "      <th>atr_std</th>\n",
       "      <th>jump_power</th>\n",
       "      <th>diff_days</th>\n",
       "      <th>wave_score1</th>\n",
       "      <th>wave_score2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>-1</td>\n",
       "      <td>0.139976</td>\n",
       "      <td>0.070454</td>\n",
       "      <td>0.068291</td>\n",
       "      <td>0.176934</td>\n",
       "      <td>-1.390000</td>\n",
       "      <td>48</td>\n",
       "      <td>0.338000</td>\n",
       "      <td>0.376933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>-1</td>\n",
       "      <td>-0.409467</td>\n",
       "      <td>-0.388299</td>\n",
       "      <td>0.750276</td>\n",
       "      <td>0.580693</td>\n",
       "      <td>-2.144082</td>\n",
       "      <td>106</td>\n",
       "      <td>0.769249</td>\n",
       "      <td>0.844728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>-1</td>\n",
       "      <td>11.841444</td>\n",
       "      <td>-0.846906</td>\n",
       "      <td>2.738821</td>\n",
       "      <td>1.438411</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0.242423</td>\n",
       "      <td>0.074236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>-1</td>\n",
       "      <td>-17.266100</td>\n",
       "      <td>-3.578966</td>\n",
       "      <td>-3.861270</td>\n",
       "      <td>2.061893</td>\n",
       "      <td>-2.621604</td>\n",
       "      <td>100</td>\n",
       "      <td>0.930900</td>\n",
       "      <td>0.966207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>1</td>\n",
       "      <td>-9.146706</td>\n",
       "      <td>4.145140</td>\n",
       "      <td>2.854837</td>\n",
       "      <td>1.022388</td>\n",
       "      <td>-1.081679</td>\n",
       "      <td>353</td>\n",
       "      <td>1.650291</td>\n",
       "      <td>1.655261</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            result  deg_hisWindowPd  deg_windowPd  deg_60WindowPd   atr_std  \\\n",
       "2015-09-11      -1         0.139976      0.070454        0.068291  0.176934   \n",
       "2015-09-11      -1        -0.409467     -0.388299        0.750276  0.580693   \n",
       "2015-09-11      -1        11.841444     -0.846906        2.738821  1.438411   \n",
       "2015-09-11      -1       -17.266100     -3.578966       -3.861270  2.061893   \n",
       "2015-09-11       1        -9.146706      4.145140        2.854837  1.022388   \n",
       "\n",
       "            jump_power  diff_days  wave_score1  wave_score2  \n",
       "2015-09-11   -1.390000         48     0.338000     0.376933  \n",
       "2015-09-11   -2.144082        106     0.769249     0.844728  \n",
       "2015-09-11    0.000000          0     0.242423     0.074236  \n",
       "2015-09-11   -2.621604        100     0.930900     0.966207  \n",
       "2015-09-11   -1.081679        353     1.650291     1.655261  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from MlFiter import MlFiterClass\n",
    "order_has_ret = orders_pd_train_snap[(orders_pd_train_snap.result <> 0)]\n",
    "\n",
    "all_pd = order_has_ret.filter(['result', 'deg_hisWindowPd', 'deg_windowPd', 'deg_60WindowPd', 'atr_std', 'jump_power', \n",
    "                               'diff_days', 'wave_score1', 'wave_score2', 'wave_score23'])\n",
    "train_np = all_pd.as_matrix()\n",
    "y = train_np[:, 0]\n",
    "x = train_np[:, 1:]\n",
    "all_filter = MlFiterClass(x, y, all_pd)\n",
    "all_pd.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 这回使用随机森林\n",
    "all_filter.estimator.random_forest_classifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "accuracy mean: 0.505349477658\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 0.48678147,  0.50788261,  0.53019646,  0.50497211,  0.51297599,\n",
       "        0.46058695,  0.51564395,  0.50994663,  0.4942975 ,  0.53021111])"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_filter.cross_val_accuracy_score()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**0.50的准确率，这算是有所提升吗？😥**\n",
    "\n",
    "### 下面使用把数据离散化后，换种分类器测试"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>result</th>\n",
       "      <th>dh_dummies_(-inf, -20]</th>\n",
       "      <th>dh_dummies_(-20, -12]</th>\n",
       "      <th>dh_dummies_(-12, -7]</th>\n",
       "      <th>dh_dummies_(-7, -3]</th>\n",
       "      <th>dh_dummies_(-3, 0]</th>\n",
       "      <th>dh_dummies_(0, 3]</th>\n",
       "      <th>dh_dummies_(3, 7]</th>\n",
       "      <th>dh_dummies_(7, 12]</th>\n",
       "      <th>dh_dummies_(12, 20]</th>\n",
       "      <th>dh_dummies_(20, inf]</th>\n",
       "      <th>dw_dummies_(-inf, -20]</th>\n",
       "      <th>dw_dummies_(-20, -12]</th>\n",
       "      <th>dw_dummies_(-12, -7]</th>\n",
       "      <th>dw_dummies_(-7, -3]</th>\n",
       "      <th>dw_dummies_(-3, 0]</th>\n",
       "      <th>dw_dummies_(0, 3]</th>\n",
       "      <th>dw_dummies_(3, 7]</th>\n",
       "      <th>dw_dummies_(7, 12]</th>\n",
       "      <th>dw_dummies_(12, 20]</th>\n",
       "      <th>dw_dummies_(20, inf]</th>\n",
       "      <th>d60_dummies_(-inf, -20]</th>\n",
       "      <th>d60_dummies_(-20, -12]</th>\n",
       "      <th>d60_dummies_(-12, -7]</th>\n",
       "      <th>d60_dummies_(-7, -3]</th>\n",
       "      <th>d60_dummies_(-3, 0]</th>\n",
       "      <th>d60_dummies_(0, 3]</th>\n",
       "      <th>d60_dummies_(3, 7]</th>\n",
       "      <th>d60_dummies_(7, 12]</th>\n",
       "      <th>d60_dummies_(12, 20]</th>\n",
       "      <th>d60_dummies_(20, inf]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            result  dh_dummies_(-inf, -20]  dh_dummies_(-20, -12]  \\\n",
       "2015-09-11       0                     0.0                    0.0   \n",
       "2015-09-11       0                     0.0                    0.0   \n",
       "2015-09-11       0                     0.0                    0.0   \n",
       "2015-09-11       0                     0.0                    1.0   \n",
       "2015-09-11       1                     0.0                    0.0   \n",
       "\n",
       "            dh_dummies_(-12, -7]  dh_dummies_(-7, -3]  dh_dummies_(-3, 0]  \\\n",
       "2015-09-11                   0.0                  0.0                 0.0   \n",
       "2015-09-11                   0.0                  0.0                 1.0   \n",
       "2015-09-11                   0.0                  0.0                 0.0   \n",
       "2015-09-11                   0.0                  0.0                 0.0   \n",
       "2015-09-11                   1.0                  0.0                 0.0   \n",
       "\n",
       "            dh_dummies_(0, 3]  dh_dummies_(3, 7]  dh_dummies_(7, 12]  \\\n",
       "2015-09-11                1.0                0.0                 0.0   \n",
       "2015-09-11                0.0                0.0                 0.0   \n",
       "2015-09-11                0.0                0.0                 1.0   \n",
       "2015-09-11                0.0                0.0                 0.0   \n",
       "2015-09-11                0.0                0.0                 0.0   \n",
       "\n",
       "            dh_dummies_(12, 20]  dh_dummies_(20, inf]  dw_dummies_(-inf, -20]  \\\n",
       "2015-09-11                  0.0                   0.0                     0.0   \n",
       "2015-09-11                  0.0                   0.0                     0.0   \n",
       "2015-09-11                  0.0                   0.0                     0.0   \n",
       "2015-09-11                  0.0                   0.0                     0.0   \n",
       "2015-09-11                  0.0                   0.0                     0.0   \n",
       "\n",
       "            dw_dummies_(-20, -12]  dw_dummies_(-12, -7]  dw_dummies_(-7, -3]  \\\n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "2015-09-11                    0.0                   0.0                  1.0   \n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "\n",
       "            dw_dummies_(-3, 0]  dw_dummies_(0, 3]  dw_dummies_(3, 7]  \\\n",
       "2015-09-11                 0.0                1.0                0.0   \n",
       "2015-09-11                 1.0                0.0                0.0   \n",
       "2015-09-11                 1.0                0.0                0.0   \n",
       "2015-09-11                 0.0                0.0                0.0   \n",
       "2015-09-11                 0.0                0.0                1.0   \n",
       "\n",
       "            dw_dummies_(7, 12]  dw_dummies_(12, 20]  dw_dummies_(20, inf]  \\\n",
       "2015-09-11                 0.0                  0.0                   0.0   \n",
       "2015-09-11                 0.0                  0.0                   0.0   \n",
       "2015-09-11                 0.0                  0.0                   0.0   \n",
       "2015-09-11                 0.0                  0.0                   0.0   \n",
       "2015-09-11                 0.0                  0.0                   0.0   \n",
       "\n",
       "            d60_dummies_(-inf, -20]  d60_dummies_(-20, -12]  \\\n",
       "2015-09-11                      0.0                     0.0   \n",
       "2015-09-11                      0.0                     0.0   \n",
       "2015-09-11                      0.0                     0.0   \n",
       "2015-09-11                      0.0                     0.0   \n",
       "2015-09-11                      0.0                     0.0   \n",
       "\n",
       "            d60_dummies_(-12, -7]  d60_dummies_(-7, -3]  d60_dummies_(-3, 0]  \\\n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "2015-09-11                    0.0                   1.0                  0.0   \n",
       "2015-09-11                    0.0                   0.0                  0.0   \n",
       "\n",
       "            d60_dummies_(0, 3]  d60_dummies_(3, 7]  d60_dummies_(7, 12]  \\\n",
       "2015-09-11                 1.0                 0.0                  0.0   \n",
       "2015-09-11                 1.0                 0.0                  0.0   \n",
       "2015-09-11                 1.0                 0.0                  0.0   \n",
       "2015-09-11                 0.0                 0.0                  0.0   \n",
       "2015-09-11                 1.0                 0.0                  0.0   \n",
       "\n",
       "            d60_dummies_(12, 20]  d60_dummies_(20, inf]  \n",
       "2015-09-11                   0.0                    0.0  \n",
       "2015-09-11                   0.0                    0.0  \n",
       "2015-09-11                   0.0                    0.0  \n",
       "2015-09-11                   0.0                    0.0  \n",
       "2015-09-11                   0.0                    0.0  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deg = MlFiterDegPdClass(orderPd=orders_pd_train_snap, dummies=True, invoke_hmm=False,\n",
    "                       invoke_pca=False)\n",
    "deg.df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**离散值的选择通过统计挖掘**\n",
    "\n",
    "上边封住好了的，在离散值的选择上不是瞎来的，会通过可视化来选择离散值bins，如下所示，这里的核心思想就是不要拍脑袋使用自己觉的对的任何常数值，**通过数据来指导数据最终指导交易**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAECCAYAAAARlssoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12623e550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "lowBkCnt show hist and qcuts\n",
      "[1, 3.3]        43098\n",
      "(3.3, 5.6]        743\n",
      "(5.6, 7.9]        243\n",
      "(7.9, 10.2]       175\n",
      "(10.2, 12.5]       42\n",
      "(12.5, 14.8]       41\n",
      "(14.8, 17.1]       40\n",
      "(17.1, 19.4]       13\n",
      "(19.4, 21.7]        9\n",
      "(21.7, 25]          8\n",
      "Name: lowBkCnt, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAECCAYAAAAFL5eMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFdhDqcBu/gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x131c33810>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "atr_std show hist and qcuts\n",
      "(1.116, 7.141]       4442\n",
      "[-0.999, -0.0925]    4442\n",
      "(0.826, 1.116]       4441\n",
      "(0.633, 0.826]       4441\n",
      "(0.483, 0.633]       4441\n",
      "(0.36, 0.483]        4441\n",
      "(0.249, 0.36]        4441\n",
      "(0.152, 0.249]       4441\n",
      "(0.0476, 0.152]      4441\n",
      "(-0.0925, 0.0476]    4441\n",
      "Name: atr_std, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAECCAYAAAASDQdFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x125905c10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "jump_power show hist and qcuts\n",
      "(-1.0792, 0]          10760\n",
      "(-1.42, -1.282]        3754\n",
      "[-151.515, -4.253]     3754\n",
      "(-1.824, -1.589]       3746\n",
      "(-2.785, -2.176]       3744\n",
      "(-1.282, -1.173]       3741\n",
      "(-1.173, -1.0792]      3738\n",
      "(-2.176, -1.824]       3734\n",
      "(-4.253, -2.785]       3725\n",
      "(-1.589, -1.42]        3716\n",
      "Name: jump_power, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAECCAYAAADkaECYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1296ee810>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "diff_days show hist and qcuts\n",
      "[0, 36.8]         17330\n",
      "(36.8, 73.6]       6851\n",
      "(73.6, 110.4]      4890\n",
      "(110.4, 147.2]     3937\n",
      "(147.2, 184]       3093\n",
      "(184, 220.8]       2256\n",
      "(220.8, 257.6]     2082\n",
      "(257.6, 294.4]     1630\n",
      "(294.4, 331.2]     1319\n",
      "(331.2, 368]       1024\n",
      "Name: diff_days, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAECCAYAAAAFL5eMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nuIuIVCAVdxGRCvT/ATV6p2cv+w77AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1296ee6d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "wave_score1 show hist and qcuts\n",
      "(1.199, 6.0266]      4442\n",
      "[-1.0169, -0.184]    4442\n",
      "(0.899, 1.199]       4441\n",
      "(0.689, 0.899]       4441\n",
      "(0.506, 0.689]       4441\n",
      "(0.353, 0.506]       4441\n",
      "(0.219, 0.353]       4441\n",
      "(0.0986, 0.219]      4441\n",
      "(-0.0262, 0.0986]    4441\n",
      "(-0.184, -0.0262]    4441\n",
      "Name: wave_score1, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAECCAYAAAAFL5eMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XuYXHWd5/F3p+nudLo7CZgmmjUhEOUHzog8uOttWQKO\nrrI647jrs+uzy/ooSmaUYUdndWZEozsrAZ+ZEZF1ZVcIF2fGCwEBx0wWyBAIiYQ0lxgu4Zum+lJJ\nd9L3e6cvqTr7R12oTqq7Kt1VdU6d/ryeh4eqc06d/qYvn/rV7/zO71fheR4iIhIuS/wuQERECk/h\nLiISQgp3EZEQUriLiISQwl1EJIQU7iIiIXRWrgOcc0uAOwEHxIE/BiaBe5PPXzaz65PHXgdsAqaB\nLWa2vThli4jIXPJpuf8+4JnZ5cBm4GbgVuBGM9sILHHOfcI5txq4AXg/8FHgFudcVZHqFhGROeQM\ndzN7hERrHOA8YAC4zMyeTm7bAXwYeA+wx8xOmtkw0AxcUviSRUQkl7z63M0s7py7F7gd+ClQkbF7\nBFgONABDGdtHgRWFKVNERM5E3hdUzeyzwIXAXUBtxq4GYBAYJhHyp24XEZESy+eC6jXAW83su8AE\nEAOec85tNLOngKuBJ4AmYItzrppE+F8EvDzXuT3P8yoqKuY6RERETpczOCtyTRzmnFsG3AO8mcSb\nwS3AayRa8FXAIeA6M/Occ58H/ij5hbeY2cM5vr7X0zOSq0bfNTY2oDoLpxzqLIcaQXUWWhnVmTPc\nc7bczWwc+E9Zdl2Z5ditwNZ8ihMRkeLRTUwiIiGkcBcRCSGFu4hICCncRURCSOEuIhJCCncRkRBS\nuIuIhJDCXUQkhBTuIiIhpHAvY5FIM5FIs99liEgAKdxFREJI4S4iEkIKdxGREFK4i4iEkMJdRCSE\nFO4iIiGkcBcRCSGFu4hICCncRURCSOEuIhJCCncRkRBSuIuIhJDCXUQkhBTuIiIhpHAXEQkhhXuZ\nikSaiUbb/S5DRAJK4S4iEkJn+V2AnBmtvCQi+VDLXUQkhOZsuTvnzgLuBtYD1cAW4Ajwa+Bw8rA7\nzGybc+46YBMwDWwxs+3FKloSOjs7Adiw4e0+VyIiQZOrW+YaoNfMPuOcOxs4APwV8D0z+37qIOfc\nauAG4DJgGbDHOfeYmU0XqW4REZlDrnC/H9iWfLyERKv83cBFzrk/JNF6/wrwHmCPmZ0Ehp1zzcAl\nwPNFqVpEROY0Z5+7mY2b2ZhzroFEyH8T2A981cw2Ai3At4HlwFDGS0eBFcUpWUREcsl5QdU5txZ4\nArjPzH4OPGxmLyZ3PwxcSiLYl2e8rAEYLHCtIiKSp1wXVFcDjwLXm9mu5OZHnXN/YmbPAb9Houul\nCdjinKsGaoGLgJfzKaCxsWG+tZdUUOocGKhPP25oWMqKFctm1BaUOnMphzrLoUZQnYVWLnXmkqvP\n/evASmCzc+5bgEeij/0259wUcBzYZGajzrnbgT1ABXCjmU3lU0BPz8i8iy+VxsaGwNTZ3z+avjN1\nZGSCoaHxdG1BqnMu5VBnOdQIqrPQyqnOXOYMdzP7MvDlLLsuz3LsVmBrvsXJ/ESj7XR2drJmzZoZ\n2yORZgYG6jn77Lf4VJmIBIluYhIRCSGFu4hICCncQ0AzRIrIqTRxWJnq7Oyku7sr/Rjg0kt/x8+S\nRCRA1HIvc/F4nO7uLuLxuN+liEiAKNzLXG9vDz/b8QK9vT1+lyIiAaJwLyOHD7/GgQMvEo/Hicfj\n9PX14Xke1bXhuOlCRApH4V5GOjqOplvpvb09PLa/nf7+fr/LEpEAUriXmcxWulrsIjIbhbuISAgp\n3EVEQkjhXmY8L05PTw99fb2n7YvFYkQizcRiMR8qE5EgUbiXmemJUXY+d4Ttu189bV80GmXT5rto\na2vxoTIRCRLdoVqGEhdSvaz7ltafU9piRCSQ1HIXEQkhhbuISAgp3EVEQkjhLiISQgr3Mjc4OOh3\nCSISQAp3EZEQUriHgOclZoiMx+NMjPZpVSYR0Tj3cjR1YnjG8+mJUZ58qZJ1617wqSIRCRq13MtI\najm9bGrqVpy2LRJpJhJpLmZJIhJQCvcy5nlxRkaGs+6LxWJEo+3EYlp+T2QxUrgHXKr1HYk0pxfE\nTpmeGOVg60jW17W1tbD5tm10dBwpRZkiEjAK9zJXtbRu1n3ZumpEZHFQuJeJWCxGX1+f32WISJlQ\nuJeJjo6jbN/9it9liEiZ0FDIMjJXFwxAb29iAY+jR4/OObJGRMJP4V4GotH2nGHteXEGBgaAWnp7\ne3jwydfV5y6yiM0Z7s65s4C7gfVANbAFeBW4F4gDL5vZ9cljrwM2AdPAFjPbXrSqF4EzHZ8+NT7M\nzmePcc5bLwZ0MVVkscvV534N0GtmVwAfBX4I3ArcaGYbgSXOuU8451YDNwDvTx53i3Ouqoh1S9L0\nxCiQuGs1V7eNiCweubpl7ge2JR9XAieBy8zs6eS2HcC/JdGK32NmJ4Fh51wzcAnwfOFLXjxisRgd\nHUfP+EYkjaoRkTlb7mY2bmZjzrkGEiH/DaAi45ARYDnQAAxlbB8F1C+wQB0dR/nm93/BSy/9lnhc\nd5qKSP5yXlB1zq0Ffgn80Mx+7pz764zdDcAgMEwi5E/dnlNjY0P+1fqo1HUODNQzNLSMiooKHnzy\ndT7zsZq8X7tsWTUwAcCKFctOq/3w4cMAXHjhhQWr90yVw8+9HGoE1Vlo5VJnLrkuqK4GHgWuN7Nd\nyc0vOueuMLPdwNXAE0ATsMU5Vw3UAhcBL+dTQE9P9tvng6SxsaHkdfb3jzI0NA4kLo5Go/kPbRwf\nn0o/HhoaP632/v5EP/1c/6bUBd0NG96e99fNlx/fzzNVDjWC6iy0cqozl1wt968DK4HNzrlvAR7w\np8D/Sl4wPQQ8YGaec+52YA+JbpsbzWxqtpOKiEhxzRnuZvZl4MtZdl2Z5ditwNbClCUiIguh6QcW\nMc33LhJeCvcA0xQCIjJfCvcyMTiY1+CjtMmxMzteRMJFc8uUgcmxQUaqRvM61vPiDA5mX51pNsUc\nGSMi/lDLPWSmxofZe7BjxrZYLEYk0kwsFvOpKhEpNYV7CKXmmPG8OAcOvMiuXf/Mps130dbWorVV\nRRYJhXuITY0P8+CTr9PVdZyaupXs2/cMP/3p37H5tm00NT1LNNoOJKYUTgR+7ha+RtiIlAeFe8il\npv6dHBvgjvv30tvbM+t0wG1tLekWvoiUN4X7IpLPHO9L689JP45G29VKFylTCvcQmzqRfdSM58Xp\n7u5K97vH4zE6OzvVDy8SIhoKGXCJRTjqGRsbBZYt6FyTY4P09S1J9sWPcO65qzn//PM5fryLO+7f\nO+drY7FY8oJsnMpKtQlEgk5/pQEVi8Xo7u4CEqstvXZ0bEHnO/WmptO7aLz018sm1R/f0XFkQXWI\nSGko3AOqo+MoP9vxQvp5VXVtXq+bOjGcXnov275cd7p6XnyOoZKepkQQKRMK9wCrri3MogGztci7\nu7vSwyFTJscG+M6PH6Wp6Vn1wYuUMYV7QBW7hex5cfr6+rIGeEVFBT/6xdP89rcv0traytTUVPpN\nINsbgogEjy6oBkyphh5OjQ/z2P4RLr74eNb9FRUV/OLRA/zT8/1UVi5h823bqF7WkH5D0Hw0IsGm\nlvsiNle3z9SJYaprG9Lj3mvqViTfENrp6sr+hiAiwaFwD6nZxrhnSo13j8fz71sv1HUAESkuhXvI\nJca29522PTWq5tdNvfT29pz2mkwaISNSfhTui8RsLfnM6QYKIXPyMU01LOIfhfsilE+XTcqJkR4O\nHXo1675sc89kTj6michE/KPRMgETjbbT2dk5592ihTTUFSFSPft7fGpUTcOqteltuYZDZn4aKPQn\nAxHJj1ruIZdYdm9+66mmWvhnchFVY+BFgkHhHjDxeGJOGc/zCnK+6YlR9h8eSj8fGZm9SybbhdeU\n2RbcjsVi7Nq1k8OHLb0tcwqDNx6r312klBTuAXP8eBc/2/ECLS0tZ9Q3Prf83ijmauHPVktHx1E2\n37ZtxoRik2MD3HTn43R0HEk/Vr+7SGkp3AMkEmmmu7uraGPJPc9LTh1cWNkWAcnW767RMyKlo3Bf\nRE5OjhHpzt6K97z4nF02c/G8eHKxj9NDO7M7JzV6JhKJzOvriEj+FO6LTNXSuqzbpydGOdg6ApzZ\nUMnOzk6mxoe5e/srM7peUoF/qpq6lbS2tqr1LlJkCvdFLrPFPlvw5+PUIY+pBblPNTk2wNf+9hH1\nwYsUWV7j3J1z7wW+a2ZXOecuBX4NHE7uvsPMtjnnrgM2AdPAFjPbXpSKpaASLfaxBQU7wMRo34xh\nkJNjg9TUrZzRLZPan3oj0MySIsWTM9ydc18D/iuQuhL3buB7Zvb9jGNWAzcAl5FY6HOPc+4xM5su\nfMlSaJnBPtuQx1wmxwbp7OxkzZo1hSpLRBYgn26Z14FPZjx/N/Ax59xTzrk7nXP1wHuAPWZ20syG\ngWbgksKXK6Uy21J9APF46gLq3LNJpt4o5l66T0SKIWe4m9lDwMmMTc8CXzOzjUAL8G1gOTCUccwo\ncPr4OMnLQkauzEdmkGe7mJq5zfPitLa2csf9e/Oe1z011r2r6/i8PxmIyJmZz9wyD5tZKsgfBm4H\nniIR8CkNQF5/xY2N5TE/eCnqHBiop66upmD94MWQuOO1goZVa2loWDpjX0PDUkZG+rMG+NL6c2Yc\nf8459enHQf0dCGpdp1KdhVUudeYyn3B/1Dn3J2b2HPB7wPNAE7DFOVcN1AIXAS/nc7KenpF5lFBa\njY0NJamzv3+UsbFJYGEjV4otdZPVyMjEjO2nPj9V5v7+/jc+LQTxd6BUP/OFUp2FVU515jKfoZBf\nBG5zzj0BfAC4ycy6SLTg9wA7gRvNbGoe55aAy9YXP3VimKkTwzNmsizc1AkiMh95tdzNrJ1EkGNm\nLwKXZzlmK7C1oNWJ72brg6+pW5n3OSbHBlla/6b089SwyXXrzitIjSJyOs3nHjBzzcwYdOVcu0jY\n6A5Vydupo2YOHXo164iZ1HHqmhHxj8Jd5iWxQlM7/f39Z/za1A1PIlI86pYJkFgsNu9Vk0pt6sQw\n1bXLZ2zLVnvqOM+Ll2zpQBFRyz0wIpFmmpr28+RzzbkPDpC53owyR9ZMjg3wsx0vlKIkEUHhHjhV\n1bV+l1A0qfHxUyeGOXToVa23KlJECncpCU07IFJaCncpOo2aESk9hbvM20ImOBscfGPEjNZWFSk8\nhXuAlNtoksyl+TzPSwZ99jVa55JaW1WrM4kUjoZCBkQ02l6Wd3imJjg7OTnGq531wJGsx6W6Zk4d\nPgmwb98zWp1JpMDUcpeCqagg3ZKfS6qVH49r8Q6RYlG4S0Hlmqp46sQwJyfHOHTMo7e3p0RViSw+\nCnfxRWrMezTaXnbXGkTKgfrcxReeFycSibBqVaPfpYiEksJdfJFarm/lyteSW872tR6RsFG3TABE\nIs2LcpbEVNeMiBSewj0g4vF42cwIWSiel/o3n/nYeBGZm8LdZ5FIM9FoO729Pew92OF3OSU1PTHK\nszbIkSPZx8aLyPwp3H0Wi8Xo7OzE87ycwwjDKN+x8SJyZhTuPorFYjQ17eeO+/fOa0WjsKhaWofn\nxYlG24nFdGOTSCEo3H20e/cufnDfDmrqVvhdiu8mxwa46c7H6ehQF41IISjcfaYRI29IzS8jIgun\ncBcRCSGFu8+0kMUbJkb7FuV4f5FiULhLIEyODTI5Nsjx48eSF1a1cIfIQijcJTA8L05rays33fm4\nFu4QWSCFuwTC1Inh5HwzQ7qwKlIAmjgsACbHBhkc1PssaPSQSKHkFe7OufcC3zWzq5xzG4B7gTjw\nspldnzzmOmATMA1sMbPtxSlZFotYLEZbWwvr119AZWWl3+WIlJWczUXn3NeAO4Ga5KZbgRvNbCOw\nxDn3CefcauAG4P3AR4FbnHNVRao5NDQy5HSZo4e0cLbI/OXTF/A68MmM5+82s6eTj3cAHwbeA+wx\ns5NmNgw0A5cUtFJZlNT/LjI/OcPdzB4CTmZsqsh4PAIsBxqAoYzto4DuqZczMj0x6ncJIqExnwuq\nmTM7NQCDwDCJkD91e06NjeVxAa0YdTY0LE0/rq1VL1amc86pBxI3Ng0N9dDYeFnJa1jMv5vFoDpL\naz7h/oJz7goz2w1cDTwBNAFbnHPVQC1wEfByPifr6Qn+dK+NjQ1FqXNkZCL9+MSJ6YKfv1xNjPZx\n4MArrFt3HgBDQ+Ml/z0p1s+80FRnYZVTnbnMZ/zdV4H/6ZzbC1QBD5hZF3A7sAfYSeKC69Q8zr0o\neV6ckRFNQ5DieXE6Ozs1/a/IAuTVcjezduADycfNwJVZjtkKbC1kcWHX3d0FJPqaD7aOLcrFOrKZ\nGh/mjvv3+l2GSFnTnTMBoWCf6dQ57iORZiKRZp+qESk/CncpC7FYTBOKiZwBhbuUhY6Oo5pQTOQM\nKNylbOiGJpH8Kdx9EovF6Ovr87uMQJptAZM3FtFW14xILgp3n7S1tfDwzia/yygrqUW01TUjkpvC\n3Qepi4NV1bV+lxJ4k2ODMyZYU9eMSH4U7j5oa2th823b/C4j0CZG+zl06FW/yxApWwp333h+FxBo\n0xOjPPnSAJPjQxw69KruVhU5Qwp3CayauhWcnBzjmRaPrq7jfpcjUlYU7hJ4nhfjmWd+43cZImVF\n4S6BNTmWmDV6anyYvQc70s9FJDeFu09mG8st2WnuHZEzo3AXEQkhhbsEXupTjj7tiORP4V5ikUgz\n+/Y943cZZUFhLjJ/CncRkRBSuPsgtQKTiEixKNxLTLNBFlYsFiMSadZMkSKnULiXWFPTfrbvfsXv\nMkKjra2FTZvv0kyRIqfIa4FsKSyN2Z4/z4uzb98ztLa2cv755wOaKVIkG4V7iaS6DtTfvjCTYwPc\ncX8bf/Sp99PVdZzVq9/sd0kigaRumRLq6DjKTx7a7XcZZa+mbgW9vT3ccf9eTSgmMguFe4mpS6Zw\naupW+F2CSGAp3KXseF5cI45EclCfewllLhcn8zfad5TH+ipoWLXW71JEAkstdykrqWl/q2sb8Ly4\nLlCLzELhXkIKosIa7TvKg0++zsRoH9Fou9/liASKwl3KyqmTiaUuqnZ2dhKJNPtRkkggzbvP3Tn3\nPDCUfNoK3AzcC8SBl83s+gVXFxKxWIxotB3P06LYIlIa82q5O+dqAMzsg8n/Pg/cCtxoZhuBJc65\nTxSwzrLW1tbC5tu20d/f73cpoaOl90Sym2/L/V1AnXPuUaAS+AZwmZk9ndy/A/gw8MjCSwwHjcku\nLM31LjK3+fa5jwN/Y2YfAb4I/ANQkbF/BFCaiYj4ZL7hfphEoGNmzUAfsDpjfwOgz8sZPC/O4KC+\nJcWQGhIZi8X9LkUkMObbLXMt8E7geufcGmA58JhzbqOZPQVcDTyRz4kaGxvmWUJpLaTOgYF6psaH\n2dtxTNMPFMHU+DAPPjnCxo19Bf19Wgy/m6WkOktrvuG+FbjHOfc0idExnyXRer/LOVcFHAIeyOdE\nPT0j8yyhdBobGxZUZ3//KKB5ZYph6sQw1bXLqalbwdDQOPv2vQDAhg1vX9B5F/ozLxXVWVjlVGcu\n8wp3M5sGrsmy68r5nC/MUsMgpbCmJ0apWlqffj45NkhnZyfr1p0HJL7vbW0trF9/AZWVlX6VKeIb\n3cRUZKlhkFI8qZEz3d1d7Nv3DNFou1ZokkVP4V4CGgbpD63QJIuZwr2I1CVTet3dXenZNz0vTjTa\nrsWzZVFSuBdRZpeMbroprtQc77FYjOPHj/Gb3+xlcmyAm+58XF0zsihpPvciq6lboVvkS2B6YpQn\nX6rk0rWDHDjiMXXiAA2r1qprRhYttdxLQK320khd26ipW0F1bUP6TTUabdeMkbLoKNwlNCZG+zl+\n/PTVrjo7O3XtQxYdhXuRjfRG/S5h0ZieGOW1o2PqBhNB4S4hU1Vd63cJIoGgcJdQmxjtSw+PVL+7\nLCYKdxGREFK4l8D0xKjfJYjIIqNwl1CbHBukr6/P7zJESk7hXiSaesBfUyeG0/cXxONazEMWH92h\nWiRtbS38xS1b/S5jUTr1prHh4WF2PneEc89dzYUXuvSF1YXO+S4SZGq5F5XndwGSVF2bWNwg9YlK\nrXgJO4W7hJrnxRkZeaMln5rM7ciRNiKRZs0YKaGlbpkCi0Sa1dceINMTozxvPaxcXU93dxfRaDs1\ndSs4ePAgN299gh9/5wvpY9VNI2GicC+CeDzGrl27/C5Dkk69azUxgmaJZoyUUFO3TIHFYjEOHjzI\n9t2v+F2KZEjN93706NEZ26LRdqamprWoh4SOWu4F1tS0n588tJuqpXV+lyIZUvO9v+lNPeltw92t\n3HTnIB//V6t48MnX+Q5w1VUf8q9IkQJSy70IFOxB5RGJRGZsSXXNaJ1bCRuFewFFIs10d3f5XYbk\n4HlxBgcHZ4yHnxwbTK+9KhIG6pYpoLa2ltNahhIciWGRo0xPxNl/uOK0/d3dXekbnAYG6jn77LeU\nukSRglHLvYCOH+9i78EOv8uQWUxPjHKwdQRI3NQ0PTHKxGif5p6RUFLLvQBisRiRSDNdXcfV3x5w\np/58Jkb7OXJkHKgHEuutxuMxhoYaWL78XNraWgCNgZfyo3AvgJ///B/4wX07GB/uZtnyRr/LkTyk\n+tsTrfkxlq1cRk9PT3qo5M1bn+Cb140DsG7deTPmo9m1ayegkTUSbAr3BXrjIqqnJd7KVNXSOqYn\nRtmxp4vdrwzxH658G54X4/HHHwM8Vq1q5JJLLmXdunXJKQviVFbO3aOpycnEbwr3eUp1xUSjUTxP\nE4SFQdXSOjwvxoEDLzI1Xs+OPccAWLZyEM/zOHask3t3HOLaj/0Ob37zag4ffi09AVllZSUbNryN\nysrKGeeMxWK0tbWwfv0Fp+0TKaaChrtzrgL4EfAuYAL4gpm1FPJrBMXu3bv4i1u2Ul27nCvfebbf\n5UiBpLppqpbG0/3z0xMj/P0je6haWk/j+kuJx+M89NBD7D88RHXtcgBq6lZyx//4HJWVlaxff0H6\nfG1tLWzafBc//s4X1IqXkir0aJk/BGrM7APA14FbC3z+wDhw4EWqaxvwvBhHjkT9LkcKKNtF8aql\ndcmfdxyz15Kjojxq6lZQU7eCpfXn0NS0n02b70pfhI1G24lG27POYROJNGvBbimqQnfLXA78PwAz\ne9Y59y8LfH7fxGIxdu/exfT0NJWVS+jr6wUyW3oaJRMW0xOjVC2tP21bde1yhrtb2W7H0j/vkd4o\n1bXLqak7m+7uLpbWn0NbWwudnceIx+MsWbIEz4vT1tZCLBZn/frzOXIkNZ+8RyTSnG7pt7W1sHbt\neRw50j5j22xdOpFIM729tQwNnVhQt4+6jsKp0OG+HBjKeH7SObfEzMpiZYSpqSmWLFlCRUVFslWV\nuNGlqambPXv28cCuw1x1yZvYvvsVpqfG0yNjFOyLQ2qETebPO7X4+XB3Kz9rH2HVeZewa9cuHt7Z\nRFX1MgDqzn4LX/2r33BWTR3vWLOE1wfq2Pi7Z7NkSQXPtMBnr76YZ5/dx4EjHle+82x2HezjU1dd\nyKpVjdzzT6/wjS98KN23v3bteVRWVhKLxXjhhedw7gL+/HuPsHnTR7jiiqvS+1K/vxs2vG3GcM5s\nQZ7qOkp1K6XeYNauPS/52oqs1xNOFYY3idT3rpz/DSmFDvdhoCHjedkEO8C2h35FV2eU1eeu4pbb\n72XFuRfwpU9fwdjYIPf98ikqKir49VPHqKhIhP70xNiM109PnSD1hjDf7TpHeZ4j9bvQ236Qhw92\nQcUbx06dSNw4dWKkh+df86iqqeNX/xzhrOpl1J/zL9i58zGef+04y1as5siREcYHR7jvl8e4/F1v\nZbRviL+85W5OjPRQ29CY/v9H3nc+j+5r5br/+K8Z7TvKV751G5e/6wE2bNjAueeu5rZ7tjM9OcbH\nN/4uF1/8DoD0OgPfvPXn3PRnn2bduvNmbN++/R954AnjUx90PPCE8aVPXzHj72DNmjXp12QTjbbP\nOPe+fc8A8L73vZ+BgXr6+0dnfW0+UnXOVcNCvfBCDzd8++4Z359C1BWNtrNu3Xklve5SUciRHs65\nfw983Myudc69D9hsZh8r2BcQEZG8FLrl/hDwYefc3uTzzxX4/CIikoeCttxFRCQYNHGYiEgIKdxF\nREJI4S4iEkIKdxGREPJ14jDn3HLg70nc/FQF/Hcz2+dnTXNxzn0S+JSZ/Re/a0kpt/l8nHPvBb5r\nZlf5XUs2zrmzgLuB9UA1sMXM/tHXorJwzi0B7gQcEAf+2Mxe9beq7Jxz5wLPAR8ys8N+1zMb59zz\nvHETZquZfd7PerJxzv0l8Ack8vJHZnbPbMf63XL/M2CnmV1JYtjk//a3nNk5524DtpDtThZ/lc18\nPs65r5EIpBq/a5nDNUCvmV0BXA380Od6ZvP7gGdmlwObgZt9rier5Jvl/wHG/a5lLs65GgAz+2Dy\nvyAG+0bg/cm/9SuBtXMd73e43wr83+TjKuCEj7Xkshf4ot9FZDFjPh8gyPP5vA580u8icrifRFhC\n4u9j2sdaZmVmjwCbkk/XAwP+VTOnvwXuAIK++vi7gDrn3KPOuZ3JT5hB8xHgZefcw8CvgF/PdXDJ\numWcc9cCXwE8Eq1fD/icmT3vnHsz8HfAfytVPbOZo85tyXfOoCmb+XzM7CHnXPHuHS8AMxsHcM41\nANuAb/hb0ezMLO6cu5fEp7dP+VzOaZxznwW6zexx59yNfteTwzjwN2a21Tn3dmCHc+7CgP0drQLW\nAR8HLiAR8BfNdnDJwt3M7ibRlzmDc+6dwE9J9LfvKVU9s5mtzgAr6/l8gsg5txb4JfBDM/uF3/XM\nxcw+m+zT3u+cu9jMgvTp93NA3Dn3YeBS4CfOuT8ws26f68rmMIlPlphZs3OuD3gLEKQV7/uAQ2Z2\nEjjsnJtwzq0ys95sB/vaLeOceweJj8H/2cwe87OWMrYX+HcAyfl8XvK3nLwE7bpFmnNuNfAo8Odm\ndp/f9cyJ7Tp5AAAA00lEQVTGOXdN8uIaJC6kx0hcWA0MM9toZlclL54fAD4T0GAHuBb4HoBzbg2J\nBtMxXys63R7go5CucRmJwM/K72X2biZxce0HyVEfg2YW9D7ZoCnH+XyCPOfF14GVwGbn3LdI1Hq1\nmU36W9Zpfgnc45x7isTf8Z8GsMZMQf6ZA2wl8f18msSb5LVB+wRsZtudc//GObefRAPpS2Y26/dV\nc8uIiISQ36NlRESkCBTuIiIhpHAXEQkhhbuISAgp3EVEQkjhLiISQgp3EZEQUriLiITQ/wcTdp8O\nFmtitwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1258aa310>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "wave_score2 show hist and qcuts\n",
      "(1.271, 5.558]       4442\n",
      "[-1.122, -0.308]     4442\n",
      "(0.994, 1.271]       4441\n",
      "(0.768, 0.994]       4441\n",
      "(0.557, 0.768]       4441\n",
      "(0.361, 0.557]       4441\n",
      "(0.172, 0.361]       4441\n",
      "(0.00248, 0.172]     4441\n",
      "(-0.138, 0.00248]    4441\n",
      "(-0.308, -0.138]     4441\n",
      "Name: wave_score2, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAECCAYAAAAFL5eMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRps/AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1266ed9d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "wave_score3 show hist and qcuts\n",
      "(1.26, 4.992]       4442\n",
      "[-1.126, -0.394]    4442\n",
      "(1.0024, 1.26]      4441\n",
      "(0.786, 1.0024]     4441\n",
      "(0.556, 0.786]      4441\n",
      "(0.322, 0.556]      4441\n",
      "(0.0828, 0.322]     4441\n",
      "(-0.107, 0.0828]    4441\n",
      "(-0.24, -0.107]     4441\n",
      "(-0.394, -0.24]     4441\n",
      "Name: wave_score3, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAECCAYAAADw0Rw8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1261a3490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "deg_60WindowPd show hist and qcuts\n",
      "(13.0319, 70.936]     4442\n",
      "[-57.515, -6.0276]    4442\n",
      "(9.102, 13.0319]      4441\n",
      "(6.817, 9.102]        4441\n",
      "(4.999, 6.817]        4441\n",
      "(3.406, 4.999]        4441\n",
      "(1.735, 3.406]        4441\n",
      "(0.1, 1.735]          4441\n",
      "(-1.955, 0.1]         4441\n",
      "(-6.0276, -1.955]     4441\n",
      "Name: deg_60WindowPd, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAECCAYAAADw0Rw8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1323c0c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "deg_hisWindowPd show hist and qcuts\n",
      "(19.339, 66.766]       4442\n",
      "[-62.804, -19.0819]    4442\n",
      "(13.183, 19.339]       4441\n",
      "(9.0385, 13.183]       4441\n",
      "(5.578, 9.0385]        4441\n",
      "(2.382, 5.578]         4441\n",
      "(-0.625, 2.382]        4441\n",
      "(-4.6, -0.625]         4441\n",
      "(-10.187, -4.6]        4441\n",
      "(-19.0819, -10.187]    4441\n",
      "Name: deg_hisWindowPd, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAECCAYAAADw0Rw8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1259b2690>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "deg_windowPd show hist and qcuts\n",
      "(8.921, 59.0546]     4442\n",
      "[-34.752, -3.519]    4442\n",
      "(6.363, 8.921]       4441\n",
      "(4.863, 6.363]       4441\n",
      "(3.776, 4.863]       4441\n",
      "(2.826, 3.776]       4441\n",
      "(1.876, 2.826]       4441\n",
      "(0.763, 1.876]       4441\n",
      "(-0.802, 0.763]      4441\n",
      "(-3.519, -0.802]     4441\n",
      "Name: deg_windowPd, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "import MlFiterBinsCs\n",
    "MlFiterBinsCs.show_orders_hist(orders_pd_train_snap)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deg().estimator.random_forest_classifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "accuracy mean: 0.487590577522\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 0.48217317,  0.51564395,  0.4962406 ,  0.49890856,  0.49745331,\n",
       "        0.48338588,  0.49527043,  0.45414847,  0.44479495,  0.50788644])"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deg().cross_val_accuracy_score()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**结果还是乱猜**\n",
    "\n",
    "用其它等等方式仍然是胡猜，自己写的类似adaboost依然效果很差\n",
    "通过非均衡概率降低假阳，提高假阴的方式，最后统计所有测试集样本的判别\n",
    "概率发现其实还瞎猜"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 提高分类数量，设置阀值，提高准确率的方案\n",
    "\n",
    "\n",
    "**通过将profit cg qcut 100份分类**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24\n",
      " 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49\n",
      " 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74\n",
      " 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>class</th>\n",
       "      <th>result</th>\n",
       "      <th>deg_hisWindowPd</th>\n",
       "      <th>deg_windowPd</th>\n",
       "      <th>deg_60WindowPd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>49</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.139976</td>\n",
       "      <td>0.070454</td>\n",
       "      <td>0.068291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>48</td>\n",
       "      <td>-1</td>\n",
       "      <td>-0.409467</td>\n",
       "      <td>-0.388299</td>\n",
       "      <td>0.750276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>48</td>\n",
       "      <td>-1</td>\n",
       "      <td>11.841444</td>\n",
       "      <td>-0.846906</td>\n",
       "      <td>2.738821</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>19</td>\n",
       "      <td>-1</td>\n",
       "      <td>-17.266100</td>\n",
       "      <td>-3.578966</td>\n",
       "      <td>-3.861270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-09-11</th>\n",
       "      <td>82</td>\n",
       "      <td>1</td>\n",
       "      <td>-9.146706</td>\n",
       "      <td>4.145140</td>\n",
       "      <td>2.854837</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            class  result  deg_hisWindowPd  deg_windowPd  deg_60WindowPd\n",
       "2015-09-11     49      -1         0.139976      0.070454        0.068291\n",
       "2015-09-11     48      -1        -0.409467     -0.388299        0.750276\n",
       "2015-09-11     48      -1        11.841444     -0.846906        2.738821\n",
       "2015-09-11     19      -1       -17.266100     -3.578966       -3.861270\n",
       "2015-09-11     82       1        -9.146706      4.145140        2.854837"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orders_pd_train_snap.profit_cg.fillna(0, inplace=True)\n",
    "order_has_ret = orders_pd_train_snap[(orders_pd_train_snap.result <> 0)]\n",
    "n_class = 100\n",
    "order_has_ret['class'] = pd.qcut(order_has_ret.profit_cg, n_class, labels=range(0, n_class))\n",
    "order_has_ret['class'] = order_has_ret['class'].astype(int)\n",
    "ZLog.info(np.unique(order_has_ret['class']))\n",
    "\n",
    "deg_pd = order_has_ret.filter(['class', 'result', 'deg_hisWindowPd', 'deg_windowPd', 'deg_60WindowPd'])\n",
    "train_np = deg_pd.as_matrix()\n",
    "y = train_np[:, 0]\n",
    "x = train_np[:, 1:]\n",
    "deg = MlFiterClass(x, y, deg_pd, force_clf=True)\n",
    "deg_pd.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* y为class train_test_split进行分类\n",
    "* 使用cv数据集的 predict < class阀值 使用result==0的情况查看概率"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>class</th>\n",
       "      <th>result</th>\n",
       "      <th>deg_hisWindowPd</th>\n",
       "      <th>deg_windowPd</th>\n",
       "      <th>deg_60WindowPd</th>\n",
       "      <th>predict</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2011-09-14</th>\n",
       "      <td>18</td>\n",
       "      <td>-1</td>\n",
       "      <td>5.985684</td>\n",
       "      <td>-6.237572</td>\n",
       "      <td>-6.587850</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-12-01</th>\n",
       "      <td>50</td>\n",
       "      <td>1</td>\n",
       "      <td>-2.001929</td>\n",
       "      <td>3.420615</td>\n",
       "      <td>5.567589</td>\n",
       "      <td>70.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-01-30</th>\n",
       "      <td>82</td>\n",
       "      <td>1</td>\n",
       "      <td>22.816944</td>\n",
       "      <td>4.173005</td>\n",
       "      <td>6.782197</td>\n",
       "      <td>87.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-11-16</th>\n",
       "      <td>3</td>\n",
       "      <td>-1</td>\n",
       "      <td>12.867986</td>\n",
       "      <td>-3.698409</td>\n",
       "      <td>-12.211724</td>\n",
       "      <td>21.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-12-18</th>\n",
       "      <td>40</td>\n",
       "      <td>-1</td>\n",
       "      <td>-12.460329</td>\n",
       "      <td>5.803071</td>\n",
       "      <td>11.280990</td>\n",
       "      <td>23.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            class  result  deg_hisWindowPd  deg_windowPd  deg_60WindowPd  \\\n",
       "2011-09-14     18      -1         5.985684     -6.237572       -6.587850   \n",
       "2011-12-01     50       1        -2.001929      3.420615        5.567589   \n",
       "2014-01-30     82       1        22.816944      4.173005        6.782197   \n",
       "2015-11-16      3      -1        12.867986     -3.698409      -12.211724   \n",
       "2015-12-18     40      -1       -12.460329      5.803071       11.280990   \n",
       "\n",
       "            predict  \n",
       "2011-09-14     18.0  \n",
       "2011-12-01     70.0  \n",
       "2014-01-30     87.0  \n",
       "2015-11-16     21.0  \n",
       "2015-12-18     23.0  "
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cross_validation import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "deg.estimator.random_forest_classifier()\n",
    "\n",
    "train_df, cv_df = train_test_split(deg.df, test_size=0.1, random_state=0)\n",
    "\n",
    "fiter = deg.get_fiter()\n",
    "fiter.fit(train_df.as_matrix()[:, 2:], train_df.as_matrix()[:, 0])\n",
    "\n",
    "predictions = fiter.predict(cv_df.as_matrix()[:, 2:])\n",
    "\n",
    "cv_df['predict'] = predictions\n",
    "cv_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0    56\n",
       "1.0    50\n",
       "2.0    43\n",
       "3.0    47\n",
       "4.0    44\n",
       "5.0    45\n",
       "6.0    47\n",
       "7.0    30\n",
       "8.0    46\n",
       "9.0    39\n",
       "Name: predict, dtype: int64"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_df['predict'].value_counts().sort_index().head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "threshold: 0, ppv: 0.607142857143\n",
      "threshold: 1, ppv: 0.632075471698\n",
      "threshold: 2, ppv: 0.58389261745\n",
      "threshold: 3, ppv: 0.586734693878\n",
      "threshold: 4, ppv: 0.579166666667\n",
      "threshold: 5, ppv: 0.571929824561\n",
      "threshold: 6, ppv: 0.551204819277\n",
      "threshold: 7, ppv: 0.541436464088\n",
      "threshold: 8, ppv: 0.526960784314\n",
      "threshold: 9, ppv: 0.523489932886\n"
     ]
    }
   ],
   "source": [
    "for threshold in np.arange(0, 10):\n",
    "    ppv = cv_df[cv_df['predict'] <= threshold].result.value_counts()\n",
    "    ZLog.info('threshold: {}, ppv: {}'.format(threshold, float(ppv[-1])/ppv.sum()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### threshold: 0，1，达到0.6了 但是数量上只有100多个，之后会有使用gmm－hmm的方式继续扩展这种思路，提高收益\n",
    "\n",
    "### 只能说最后一种的这个思想可以作为扩展使用，实际中由于 '非均衡胜负收益' 很难带来质的好转，突然想到阿西莫夫的小说基地，对心理史学的最开始的评估，‘可能但是不可行’，我们的印钞机之路是否也是这样呢，下一章开始就是真正的印钞机之路了，一个即可能又可行之路！！！ 下一章开始将使用深度学习方法来训练数据，提高胜率主要是卷机神经网络的使用"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 感谢🙏您能有耐心看到这里\n",
    "## 如果有什么问题可以加阿布的微信 \n",
    "## 微信号：aaaabbbuu｀"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![image](./mmexport1475383814280.jpg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}