pandas-ml-utils-torch/examples/regression_with_regularization.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from pandas_ml_utils import pd\n"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Fist lets load some data"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "source": [
    "df = pd.read_csv('SPY.csv')\n",
    "df.tail()\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "execution_count": 2,
   "outputs": [
    {
     "data": {
      "text/plain": "            Date        Open        High         Low       Close   Adj Close  \\\n6758  2019-11-29  314.859985  315.130005  314.059998  314.309998  314.309998   \n6759  2019-12-02  314.589996  314.660004  311.170013  311.640015  311.640015   \n6760  2019-12-03  308.649994  309.640015  307.130005  309.549988  309.549988   \n6761  2019-12-04  310.700012  312.119995  310.320007  311.459991  311.459991   \n6762  2019-12-05  312.230011  312.250000  310.580109  312.089996  312.089996   \n\n        Volume  \n6758  36592700  \n6759  75767800  \n6760  73941700  \n6761  49080000  \n6762  40781669  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Date</th>\n      <th>Open</th>\n      <th>High</th>\n      <th>Low</th>\n      <th>Close</th>\n      <th>Adj Close</th>\n      <th>Volume</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>6758</th>\n      <td>2019-11-29</td>\n      <td>314.859985</td>\n      <td>315.130005</td>\n      <td>314.059998</td>\n      <td>314.309998</td>\n      <td>314.309998</td>\n      <td>36592700</td>\n    </tr>\n    <tr>\n      <th>6759</th>\n      <td>2019-12-02</td>\n      <td>314.589996</td>\n      <td>314.660004</td>\n      <td>311.170013</td>\n      <td>311.640015</td>\n      <td>311.640015</td>\n      <td>75767800</td>\n    </tr>\n    <tr>\n      <th>6760</th>\n      <td>2019-12-03</td>\n      <td>308.649994</td>\n      <td>309.640015</td>\n      <td>307.130005</td>\n      <td>309.549988</td>\n      <td>309.549988</td>\n      <td>73941700</td>\n    </tr>\n    <tr>\n      <th>6761</th>\n      <td>2019-12-04</td>\n      <td>310.700012</td>\n      <td>312.119995</td>\n      <td>310.320007</td>\n      <td>311.459991</td>\n      <td>311.459991</td>\n      <td>49080000</td>\n    </tr>\n    <tr>\n      <th>6762</th>\n      <td>2019-12-05</td>\n      <td>312.230011</td>\n      <td>312.250000</td>\n      <td>310.580109</td>\n      <td>312.089996</td>\n      <td>312.089996</td>\n      <td>40781669</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Fit Regularized Model\n",
    "Now we just build a very naive model where we add an L1 and L2 penalization term. Don't expect the model to predict\n",
    "anything useful"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 0, train loss: 0.4192713499069214, test loss: 0.4193301796913147\n",
      "epoch: 1, train loss: 0.41117382049560547, test loss: 0.41122931241989136\n",
      "epoch: 2, train loss: 0.4032788872718811, test loss: 0.40333104133605957\n",
      "epoch: 3, train loss: 0.3955911099910736, test loss: 0.39563998579978943\n",
      "epoch: 4, train loss: 0.3881143033504486, test loss: 0.3881598711013794\n",
      "epoch: 5, train loss: 0.3808514177799225, test loss: 0.3808937072753906\n",
      "epoch: 6, train loss: 0.3738045394420624, test loss: 0.37384361028671265\n",
      "epoch: 7, train loss: 0.36697521805763245, test loss: 0.36701107025146484\n",
      "epoch: 8, train loss: 0.3603639006614685, test loss: 0.3603966236114502\n",
      "epoch: 9, train loss: 0.3539707064628601, test loss: 0.3540003299713135\n",
      "epoch: 10, train loss: 0.3477947413921356, test loss: 0.34782135486602783\n",
      "epoch: 11, train loss: 0.34183475375175476, test loss: 0.34185829758644104\n",
      "epoch: 12, train loss: 0.3360885977745056, test loss: 0.3361091613769531\n",
      "epoch: 13, train loss: 0.3305537700653076, test loss: 0.33057141304016113\n",
      "epoch: 14, train loss: 0.32522711157798767, test loss: 0.32524192333221436\n",
      "epoch: 15, train loss: 0.3201051950454712, test loss: 0.32011717557907104\n",
      "epoch: 16, train loss: 0.31518399715423584, test loss: 0.31519320607185364\n",
      "epoch: 17, train loss: 0.3104591369628906, test loss: 0.31046560406684875\n",
      "epoch: 18, train loss: 0.30592602491378784, test loss: 0.30592983961105347\n",
      "epoch: 19, train loss: 0.30157971382141113, test loss: 0.30158090591430664\n",
      "epoch: 20, train loss: 0.29741498827934265, test loss: 0.2974136471748352\n",
      "epoch: 21, train loss: 0.29342660307884216, test loss: 0.29342275857925415\n",
      "epoch: 22, train loss: 0.2896089255809784, test loss: 0.2896026372909546\n",
      "epoch: 23, train loss: 0.28595638275146484, test loss: 0.28594771027565\n",
      "epoch: 24, train loss: 0.282463401556015, test loss: 0.2824523448944092\n",
      "epoch: 25, train loss: 0.2791241407394409, test loss: 0.27911078929901123\n",
      "epoch: 26, train loss: 0.2759329080581665, test loss: 0.2759173512458801\n",
      "epoch: 27, train loss: 0.27288392186164856, test loss: 0.2728661894798279\n",
      "epoch: 28, train loss: 0.2699715495109558, test loss: 0.2699517011642456\n",
      "epoch: 29, train loss: 0.2671900987625122, test loss: 0.26716819405555725\n",
      "epoch: 30, train loss: 0.2645341455936432, test loss: 0.26451021432876587\n",
      "epoch: 31, train loss: 0.2619982063770294, test loss: 0.2619723081588745\n",
      "epoch: 32, train loss: 0.2595769166946411, test loss: 0.259549081325531\n",
      "epoch: 33, train loss: 0.25726503133773804, test loss: 0.2572353482246399\n",
      "epoch: 34, train loss: 0.2550576627254486, test loss: 0.2550261318683624\n",
      "epoch: 35, train loss: 0.25294971466064453, test loss: 0.2529164254665375\n",
      "epoch: 36, train loss: 0.25093644857406616, test loss: 0.250901460647583\n",
      "epoch: 37, train loss: 0.24901330471038818, test loss: 0.24897661805152893\n",
      "epoch: 38, train loss: 0.24717578291893005, test loss: 0.24713747203350067\n",
      "epoch: 39, train loss: 0.24541954696178436, test loss: 0.24537965655326843\n",
      "epoch: 40, train loss: 0.24374055862426758, test loss: 0.24369913339614868\n",
      "epoch: 41, train loss: 0.24213477969169617, test loss: 0.24209186434745789\n",
      "epoch: 42, train loss: 0.24059845507144928, test loss: 0.2405540943145752\n",
      "epoch: 43, train loss: 0.23912787437438965, test loss: 0.23908209800720215\n",
      "epoch: 44, train loss: 0.2377195656299591, test loss: 0.23767241835594177\n",
      "epoch: 45, train loss: 0.2363702356815338, test loss: 0.23632177710533142\n",
      "epoch: 46, train loss: 0.2350766807794571, test loss: 0.23502692580223083\n",
      "epoch: 47, train loss: 0.2338358759880066, test loss: 0.23378485441207886\n",
      "epoch: 48, train loss: 0.23264490067958832, test loss: 0.23259267210960388\n",
      "epoch: 49, train loss: 0.23150096833705902, test loss: 0.23144756257534027\n",
      "epoch: 50, train loss: 0.23040157556533813, test loss: 0.23034703731536865\n",
      "epoch: 51, train loss: 0.22934414446353912, test loss: 0.2292884886264801\n",
      "epoch: 52, train loss: 0.22832639515399933, test loss: 0.22826965153217316\n",
      "epoch: 53, train loss: 0.22734606266021729, test loss: 0.22728827595710754\n",
      "epoch: 54, train loss: 0.2264009714126587, test loss: 0.22634217143058777\n",
      "epoch: 55, train loss: 0.22548913955688477, test loss: 0.22542935609817505\n",
      "epoch: 56, train loss: 0.22460874915122986, test loss: 0.22454801201820374\n",
      "epoch: 57, train loss: 0.22375789284706116, test loss: 0.22369623184204102\n",
      "epoch: 58, train loss: 0.22293493151664734, test loss: 0.22287237644195557\n",
      "epoch: 59, train loss: 0.22213824093341827, test loss: 0.22207482159137726\n",
      "epoch: 60, train loss: 0.221366286277771, test loss: 0.22130203247070312\n",
      "epoch: 61, train loss: 0.22061768174171448, test loss: 0.22055259346961975\n",
      "epoch: 62, train loss: 0.2198910117149353, test loss: 0.2198251485824585\n",
      "epoch: 63, train loss: 0.21918508410453796, test loss: 0.21911844611167908\n",
      "epoch: 64, train loss: 0.2184986174106598, test loss: 0.2184312343597412\n",
      "epoch: 65, train loss: 0.2178332358598709, test loss: 0.21776515245437622\n",
      "epoch: 66, train loss: 0.21718759834766388, test loss: 0.21711879968643188\n",
      "epoch: 67, train loss: 0.21655575931072235, test loss: 0.21648629009723663\n",
      "epoch: 68, train loss: 0.2159416675567627, test loss: 0.21587154269218445\n",
      "epoch: 69, train loss: 0.21533960103988647, test loss: 0.2152688354253769\n",
      "epoch: 70, train loss: 0.21475347876548767, test loss: 0.21468210220336914\n",
      "epoch: 71, train loss: 0.21417775750160217, test loss: 0.2141057848930359\n",
      "epoch: 72, train loss: 0.21361643075942993, test loss: 0.2135438770055771\n",
      "epoch: 73, train loss: 0.2130640745162964, test loss: 0.21299096941947937\n",
      "epoch: 74, train loss: 0.21252477169036865, test loss: 0.21245113015174866\n",
      "epoch: 75, train loss: 0.21199306845664978, test loss: 0.2119188904762268\n",
      "epoch: 76, train loss: 0.2114732712507248, test loss: 0.21139858663082123\n",
      "epoch: 77, train loss: 0.21095994114875793, test loss: 0.21088474988937378\n",
      "epoch: 78, train loss: 0.21045741438865662, test loss: 0.21038177609443665\n",
      "epoch: 79, train loss: 0.20996037125587463, test loss: 0.20988425612449646\n",
      "epoch: 80, train loss: 0.20947319269180298, test loss: 0.209396630525589\n",
      "epoch: 81, train loss: 0.20899063348770142, test loss: 0.2089136391878128\n",
      "epoch: 82, train loss: 0.20851707458496094, test loss: 0.2084396779537201\n",
      "epoch: 83, train loss: 0.20804734528064728, test loss: 0.207969531416893\n",
      "epoch: 84, train loss: 0.207585871219635, test loss: 0.2075076699256897\n",
      "epoch: 85, train loss: 0.20712755620479584, test loss: 0.2070489525794983\n",
      "epoch: 86, train loss: 0.20667684078216553, test loss: 0.20659786462783813\n",
      "epoch: 87, train loss: 0.20622867345809937, test loss: 0.20614933967590332\n",
      "epoch: 88, train loss: 0.2057875245809555, test loss: 0.2057078629732132\n",
      "epoch: 89, train loss: 0.205348402261734, test loss: 0.20526841282844543\n",
      "epoch: 90, train loss: 0.2049158215522766, test loss: 0.20483550429344177\n",
      "epoch: 91, train loss: 0.2044847458600998, test loss: 0.2044041007757187\n",
      "epoch: 92, train loss: 0.20405974984169006, test loss: 0.20397880673408508\n",
      "epoch: 93, train loss: 0.20363585650920868, test loss: 0.20355461537837982\n",
      "epoch: 94, train loss: 0.20321768522262573, test loss: 0.203136146068573\n",
      "epoch: 95, train loss: 0.2028001993894577, test loss: 0.20271839201450348\n",
      "epoch: 96, train loss: 0.20238810777664185, test loss: 0.20230603218078613\n",
      "epoch: 97, train loss: 0.20197640359401703, test loss: 0.20189407467842102\n",
      "epoch: 98, train loss: 0.20156976580619812, test loss: 0.20148718357086182\n",
      "epoch: 99, train loss: 0.20116320252418518, test loss: 0.20108038187026978\n",
      "saved model to: /tmp/regression_with_regularization.model\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/kic/.local/lib/python3.8/site-packages/torch/storage.py:34: FutureWarning: pickle support for Storage will be removed in 1.5. Use `torch.save` instead\n",
      "  warnings.warn(\"pickle support for Storage will be removed in 1.5. Use `torch.save` instead\", FutureWarning)\n",
      "mean_gamma_deviance failed: (ValueError('Mean Tweedie deviance error with power=2 can only be used on strictly positive y and y_pred.'), 'error while calling <function mean_gamma_deviance at 0x7f5c7894b4c0>(FullArgSpec(args=[], varargs=\\'args\\', varkw=\\'kwargs\\', defaults=None, kwonlyargs=[], kwonlydefaults=None, annotations={}))\\n@_deprecate_positional_args\\ndef mean_gamma_deviance(y_true, y_pred, *, sample_weight=None):\\n    \"\"\"Mean Gamma deviance regression loss.\\n\\n    Gamma deviance is equivalent to the Tweedie deviance with\\n    the power parameter `power=2`. It is invariant to scaling of\\n    the target variable, and measures relative errors.\\n\\n    Read more in the :ref:`User Guide <mean_tweedie_deviance>`.\\n\\n    Parameters\\n    ----------\\n    y_true : array-like of shape (n_samples,)\\n        Ground truth (correct) target values. Requires y_true > 0.\\n\\n    y_pred : array-like of shape (n_samples,)\\n        Estimated target values. Requires y_pred > 0.\\n\\n    sample_weight : array-like of shape (n_samples,), default=None\\n        Sample weights.\\n\\n    Returns\\n    -------\\n    loss : float\\n        A non-negative floating point value (the best value is 0.0).\\n\\n    Examples\\n    --------\\n    >>> from sklearn.metrics import mean_gamma_deviance\\n    >>> y_true = [2, 0.5, 1, 4]\\n    >>> y_pred = [0.5, 0.5, 2., 2.]\\n    >>> mean_gamma_deviance(y_true, y_pred)\\n    1.0568...\\n    \"\"\"\\n    return mean_tweedie_deviance(\\n        y_true, y_pred, sample_weight=sample_weight, power=2\\n    )\\n\\nwith arguments:\\n<BoundArguments (y_true=array([[-0.02522775],\\n       [-0.00071891],\\n       [-0.00071942],\\n       ...,\\n       [ 0.01882243],\\n       [-0.00947422],\\n       [ 0.00494184]]), y_pred=array([[-0.02081658],\\n       [-0.02082147],\\n       [-0.02126519],\\n       ...,\\n       [-0.02411348],\\n       [-0.02465902],\\n       [-0.01991536]], dtype=float32), sample_weight=None)>, {\\'sample_weight\\': None}')\n",
      "mean_squared_log_error failed: (ValueError('Mean Squared Logarithmic Error cannot be used when targets contain negative values.'), 'error while calling <function mean_squared_log_error at 0x7f5c78947d30>(FullArgSpec(args=[], varargs=\\'args\\', varkw=\\'kwargs\\', defaults=None, kwonlyargs=[], kwonlydefaults=None, annotations={}))\\n@_deprecate_positional_args\\ndef mean_squared_log_error(y_true, y_pred, *,\\n                           sample_weight=None,\\n                           multioutput=\\'uniform_average\\'):\\n    \"\"\"Mean squared logarithmic error regression loss\\n\\n    Read more in the :ref:`User Guide <mean_squared_log_error>`.\\n\\n    Parameters\\n    ----------\\n    y_true : array-like of shape (n_samples,) or (n_samples, n_outputs)\\n        Ground truth (correct) target values.\\n\\n    y_pred : array-like of shape (n_samples,) or (n_samples, n_outputs)\\n        Estimated target values.\\n\\n    sample_weight : array-like of shape (n_samples,), optional\\n        Sample weights.\\n\\n    multioutput : string in [\\'raw_values\\', \\'uniform_average\\'] \\\\\\n            or array-like of shape (n_outputs)\\n\\n        Defines aggregating of multiple output values.\\n        Array-like value defines weights used to average errors.\\n\\n        \\'raw_values\\' :\\n            Returns a full set of errors when the input is of multioutput\\n            format.\\n\\n        \\'uniform_average\\' :\\n            Errors of all outputs are averaged with uniform weight.\\n\\n    Returns\\n    -------\\n    loss : float or ndarray of floats\\n        A non-negative floating point value (the best value is 0.0), or an\\n        array of floating point values, one for each individual target.\\n\\n    Examples\\n    --------\\n    >>> from sklearn.metrics import mean_squared_log_error\\n    >>> y_true = [3, 5, 2.5, 7]\\n    >>> y_pred = [2.5, 5, 4, 8]\\n    >>> mean_squared_log_error(y_true, y_pred)\\n    0.039...\\n    >>> y_true = [[0.5, 1], [1, 2], [7, 6]]\\n    >>> y_pred = [[0.5, 2], [1, 2.5], [8, 8]]\\n    >>> mean_squared_log_error(y_true, y_pred)\\n    0.044...\\n    >>> mean_squared_log_error(y_true, y_pred, multioutput=\\'raw_values\\')\\n    array([0.00462428, 0.08377444])\\n    >>> mean_squared_log_error(y_true, y_pred, multioutput=[0.3, 0.7])\\n    0.060...\\n\\n    \"\"\"\\n    y_type, y_true, y_pred, multioutput = _check_reg_targets(\\n        y_true, y_pred, multioutput)\\n    check_consistent_length(y_true, y_pred, sample_weight)\\n\\n    if (y_true < 0).any() or (y_pred < 0).any():\\n        raise ValueError(\"Mean Squared Logarithmic Error cannot be used when \"\\n                         \"targets contain negative values.\")\\n\\n    return mean_squared_error(np.log1p(y_true), np.log1p(y_pred),\\n                              sample_weight=sample_weight,\\n                              multioutput=multioutput)\\n\\nwith arguments:\\n<BoundArguments (y_true=array([[-0.02522775],\\n       [-0.00071891],\\n       [-0.00071942],\\n       ...,\\n       [ 0.01882243],\\n       [-0.00947422],\\n       [ 0.00494184]]), y_pred=array([[-0.02081658],\\n       [-0.02082147],\\n       [-0.02126519],\\n       ...,\\n       [-0.02411348],\\n       [-0.02465902],\\n       [-0.01991536]], dtype=float32), sample_weight=None)>, {\\'sample_weight\\': None}')\n",
      "mean_poisson_deviance failed: (ValueError('Mean Tweedie deviance error with power=1 can only be used on non-negative y and strictly positive y_pred.'), 'error while calling <function mean_poisson_deviance at 0x7f5c7894b3a0>(FullArgSpec(args=[], varargs=\\'args\\', varkw=\\'kwargs\\', defaults=None, kwonlyargs=[], kwonlydefaults=None, annotations={}))\\n@_deprecate_positional_args\\ndef mean_poisson_deviance(y_true, y_pred, *, sample_weight=None):\\n    \"\"\"Mean Poisson deviance regression loss.\\n\\n    Poisson deviance is equivalent to the Tweedie deviance with\\n    the power parameter `power=1`.\\n\\n    Read more in the :ref:`User Guide <mean_tweedie_deviance>`.\\n\\n    Parameters\\n    ----------\\n    y_true : array-like of shape (n_samples,)\\n        Ground truth (correct) target values. Requires y_true >= 0.\\n\\n    y_pred : array-like of shape (n_samples,)\\n        Estimated target values. Requires y_pred > 0.\\n\\n    sample_weight : array-like of shape (n_samples,), default=None\\n        Sample weights.\\n\\n    Returns\\n    -------\\n    loss : float\\n        A non-negative floating point value (the best value is 0.0).\\n\\n    Examples\\n    --------\\n    >>> from sklearn.metrics import mean_poisson_deviance\\n    >>> y_true = [2, 0, 1, 4]\\n    >>> y_pred = [0.5, 0.5, 2., 2.]\\n    >>> mean_poisson_deviance(y_true, y_pred)\\n    1.4260...\\n    \"\"\"\\n    return mean_tweedie_deviance(\\n        y_true, y_pred, sample_weight=sample_weight, power=1\\n    )\\n\\nwith arguments:\\n<BoundArguments (y_true=array([[-0.02522775],\\n       [-0.00071891],\\n       [-0.00071942],\\n       ...,\\n       [ 0.01882243],\\n       [-0.00947422],\\n       [ 0.00494184]]), y_pred=array([[-0.02081658],\\n       [-0.02082147],\\n       [-0.02126519],\\n       ...,\\n       [-0.02411348],\\n       [-0.02465902],\\n       [-0.01991536]], dtype=float32), sample_weight=None)>, {\\'sample_weight\\': None}')\n",
      "mean_gamma_deviance failed: (ValueError('Mean Tweedie deviance error with power=2 can only be used on strictly positive y and y_pred.'), 'error while calling <function mean_gamma_deviance at 0x7f5c7894b4c0>(FullArgSpec(args=[], varargs=\\'args\\', varkw=\\'kwargs\\', defaults=None, kwonlyargs=[], kwonlydefaults=None, annotations={}))\\n@_deprecate_positional_args\\ndef mean_gamma_deviance(y_true, y_pred, *, sample_weight=None):\\n    \"\"\"Mean Gamma deviance regression loss.\\n\\n    Gamma deviance is equivalent to the Tweedie deviance with\\n    the power parameter `power=2`. It is invariant to scaling of\\n    the target variable, and measures relative errors.\\n\\n    Read more in the :ref:`User Guide <mean_tweedie_deviance>`.\\n\\n    Parameters\\n    ----------\\n    y_true : array-like of shape (n_samples,)\\n        Ground truth (correct) target values. Requires y_true > 0.\\n\\n    y_pred : array-like of shape (n_samples,)\\n        Estimated target values. Requires y_pred > 0.\\n\\n    sample_weight : array-like of shape (n_samples,), default=None\\n        Sample weights.\\n\\n    Returns\\n    -------\\n    loss : float\\n        A non-negative floating point value (the best value is 0.0).\\n\\n    Examples\\n    --------\\n    >>> from sklearn.metrics import mean_gamma_deviance\\n    >>> y_true = [2, 0.5, 1, 4]\\n    >>> y_pred = [0.5, 0.5, 2., 2.]\\n    >>> mean_gamma_deviance(y_true, y_pred)\\n    1.0568...\\n    \"\"\"\\n    return mean_tweedie_deviance(\\n        y_true, y_pred, sample_weight=sample_weight, power=2\\n    )\\n\\nwith arguments:\\n<BoundArguments (y_true=array([[-0.01586294],\\n       [-0.01587686],\\n       [-0.0010646 ],\\n       ...,\\n       [-0.00670654],\\n       [ 0.00617026],\\n       [ 0.00202275]]), y_pred=array([[-0.02337915],\\n       [-0.01985289],\\n       [-0.02050145],\\n       ...,\\n       [-0.02137096],\\n       [-0.0221484 ],\\n       [-0.02201511]], dtype=float32), sample_weight=None)>, {\\'sample_weight\\': None}')\n",
      "mean_squared_log_error failed: (ValueError('Mean Squared Logarithmic Error cannot be used when targets contain negative values.'), 'error while calling <function mean_squared_log_error at 0x7f5c78947d30>(FullArgSpec(args=[], varargs=\\'args\\', varkw=\\'kwargs\\', defaults=None, kwonlyargs=[], kwonlydefaults=None, annotations={}))\\n@_deprecate_positional_args\\ndef mean_squared_log_error(y_true, y_pred, *,\\n                           sample_weight=None,\\n                           multioutput=\\'uniform_average\\'):\\n    \"\"\"Mean squared logarithmic error regression loss\\n\\n    Read more in the :ref:`User Guide <mean_squared_log_error>`.\\n\\n    Parameters\\n    ----------\\n    y_true : array-like of shape (n_samples,) or (n_samples, n_outputs)\\n        Ground truth (correct) target values.\\n\\n    y_pred : array-like of shape (n_samples,) or (n_samples, n_outputs)\\n        Estimated target values.\\n\\n    sample_weight : array-like of shape (n_samples,), optional\\n        Sample weights.\\n\\n    multioutput : string in [\\'raw_values\\', \\'uniform_average\\'] \\\\\\n            or array-like of shape (n_outputs)\\n\\n        Defines aggregating of multiple output values.\\n        Array-like value defines weights used to average errors.\\n\\n        \\'raw_values\\' :\\n            Returns a full set of errors when the input is of multioutput\\n            format.\\n\\n        \\'uniform_average\\' :\\n            Errors of all outputs are averaged with uniform weight.\\n\\n    Returns\\n    -------\\n    loss : float or ndarray of floats\\n        A non-negative floating point value (the best value is 0.0), or an\\n        array of floating point values, one for each individual target.\\n\\n    Examples\\n    --------\\n    >>> from sklearn.metrics import mean_squared_log_error\\n    >>> y_true = [3, 5, 2.5, 7]\\n    >>> y_pred = [2.5, 5, 4, 8]\\n    >>> mean_squared_log_error(y_true, y_pred)\\n    0.039...\\n    >>> y_true = [[0.5, 1], [1, 2], [7, 6]]\\n    >>> y_pred = [[0.5, 2], [1, 2.5], [8, 8]]\\n    >>> mean_squared_log_error(y_true, y_pred)\\n    0.044...\\n    >>> mean_squared_log_error(y_true, y_pred, multioutput=\\'raw_values\\')\\n    array([0.00462428, 0.08377444])\\n    >>> mean_squared_log_error(y_true, y_pred, multioutput=[0.3, 0.7])\\n    0.060...\\n\\n    \"\"\"\\n    y_type, y_true, y_pred, multioutput = _check_reg_targets(\\n        y_true, y_pred, multioutput)\\n    check_consistent_length(y_true, y_pred, sample_weight)\\n\\n    if (y_true < 0).any() or (y_pred < 0).any():\\n        raise ValueError(\"Mean Squared Logarithmic Error cannot be used when \"\\n                         \"targets contain negative values.\")\\n\\n    return mean_squared_error(np.log1p(y_true), np.log1p(y_pred),\\n                              sample_weight=sample_weight,\\n                              multioutput=multioutput)\\n\\nwith arguments:\\n<BoundArguments (y_true=array([[-0.01586294],\\n       [-0.01587686],\\n       [-0.0010646 ],\\n       ...,\\n       [-0.00670654],\\n       [ 0.00617026],\\n       [ 0.00202275]]), y_pred=array([[-0.02337915],\\n       [-0.01985289],\\n       [-0.02050145],\\n       ...,\\n       [-0.02137096],\\n       [-0.0221484 ],\\n       [-0.02201511]], dtype=float32), sample_weight=None)>, {\\'sample_weight\\': None}')\n",
      "mean_poisson_deviance failed: (ValueError('Mean Tweedie deviance error with power=1 can only be used on non-negative y and strictly positive y_pred.'), 'error while calling <function mean_poisson_deviance at 0x7f5c7894b3a0>(FullArgSpec(args=[], varargs=\\'args\\', varkw=\\'kwargs\\', defaults=None, kwonlyargs=[], kwonlydefaults=None, annotations={}))\\n@_deprecate_positional_args\\ndef mean_poisson_deviance(y_true, y_pred, *, sample_weight=None):\\n    \"\"\"Mean Poisson deviance regression loss.\\n\\n    Poisson deviance is equivalent to the Tweedie deviance with\\n    the power parameter `power=1`.\\n\\n    Read more in the :ref:`User Guide <mean_tweedie_deviance>`.\\n\\n    Parameters\\n    ----------\\n    y_true : array-like of shape (n_samples,)\\n        Ground truth (correct) target values. Requires y_true >= 0.\\n\\n    y_pred : array-like of shape (n_samples,)\\n        Estimated target values. Requires y_pred > 0.\\n\\n    sample_weight : array-like of shape (n_samples,), default=None\\n        Sample weights.\\n\\n    Returns\\n    -------\\n    loss : float\\n        A non-negative floating point value (the best value is 0.0).\\n\\n    Examples\\n    --------\\n    >>> from sklearn.metrics import mean_poisson_deviance\\n    >>> y_true = [2, 0, 1, 4]\\n    >>> y_pred = [0.5, 0.5, 2., 2.]\\n    >>> mean_poisson_deviance(y_true, y_pred)\\n    1.4260...\\n    \"\"\"\\n    return mean_tweedie_deviance(\\n        y_true, y_pred, sample_weight=sample_weight, power=1\\n    )\\n\\nwith arguments:\\n<BoundArguments (y_true=array([[-0.01586294],\\n       [-0.01587686],\\n       [-0.0010646 ],\\n       ...,\\n       [-0.00670654],\\n       [ 0.00617026],\\n       [ 0.00202275]]), y_pred=array([[-0.02337915],\\n       [-0.01985289],\\n       [-0.02050145],\\n       ...,\\n       [-0.02137096],\\n       [-0.0221484 ],\\n       [-0.02201511]], dtype=float32), sample_weight=None)>, {\\'sample_weight\\': None}')\n"
     ]
    },
    {
     "data": {
      "text/plain": "<pandas_ml_utils.ml.fitting.fit.Fit at 0x7f5c2093ce20>",
      "text/html": "<style>\n    .left {\n        float:left;\n    }\n    .right {\n        float:right;\n        text-align: right;\n    }\n    .full-width {\n        width: 100%;\n        overflow-x: scroll;\n    }\n    .no-float {\n        float: none;\n    }\n    .no-border {\n        border: none;\n    }\n    .big {\n        width: 100%;\n        max-width: 100em !important;\n        max-height: 100em;\n    }\n    .va-top {\n        vertical-align: top;\n    }\n</style>\n\n<div style=\"width: 100%; overflow-x: scroll;\">\n        <h2>Loss</h2>\n        <div class=\"full-width\">\n            <img src=\"data:image/png;base64, iVBORw0KGgoAAAANSUhEUgAAAeoAAAFlCAYAAAAki6s3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKGi9dF7LV0EAAAAASUVORK5CYII=\n\" />\n        </div>\n\n    <h2>Summary</h2>\n    <table class=\"full-width\">\n        <thead>\n            <tr>\n                <th><h3 class=\"left\">Training Data</h3></th>\n                <th><h3 class=\"right\">Test Data</h3></th>\n            </tr>\n        </thead>\n        <tbody>\n            <tr>\n                <td width=\"50%\"><style>\n    .left {\n        float:left;\n    }\n    .right {\n        float:right;\n        text-align: right;\n    }\n    .full-width {\n        width: 100%;\n        overflow-x: scroll;\n    }\n    .no-float {\n        float: none;\n    }\n    .no-border {\n        border: none;\n    }\n    .big {\n        width: 100%;\n        max-width: 100em !important;\n        max-height: 100em;\n    }\n    .va-top {\n        vertical-align: top;\n    }\n</style>\n\n<div class=\"full-width\">\n        <table width=\"100%\" class=\"full-width no-boder\" cellspacing=\"0\" cellpadding=\"0\">\n                <tr class=\"no-border\">\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"1\" width=\"50.0%\">\n                                <img src=\"data:image/png;base64, iVBORw0KGgoAAAANSUhEUgAAAYgAAAFzCAYAAAAkFp78AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEBGRaSkgRERkWv8/Q+fYm7EchOAAAAAASUVORK5CYII=\n\" class=\"left big\" />\n                        </td>\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"1\" width=\"50.0%\">\n                                <div class=\"full-width\">\n                                    <style  type=\"text/css\" >\n#T_16294728_6b08_11eb_a2c4_e991cc078694row0_col0{\n            background-color:  #a5002695;\n        }#T_16294728_6b08_11eb_a2c4_e991cc078694row1_col0{\n            background-color:  #fffdbc95;\n        }#T_16294728_6b08_11eb_a2c4_e991cc078694row2_col0{\n            background-color:  #00000095;\n        }</style><table id=\"T_16294728_6b08_11eb_a2c4_e991cc078694\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >0</th>    </tr></thead><tbody>\n                <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row0\" class=\"row_heading level0 row0\" >r2_score</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row0_col0\" class=\"data row0 col0\" >-3.14</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row1\" class=\"row_heading level0 row1\" >explained_variance_score</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row1_col0\" class=\"data row1 col0\" >-0.01</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row2\" class=\"row_heading level0 row2\" >mean_gamma_deviance</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row2_col0\" class=\"data row2 col0\" >nan</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row3\" class=\"row_heading level0 row3\" >max_error</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row3_col0\" class=\"data row3 col0\" >0.17</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row4\" class=\"row_heading level0 row4\" >mean_absolute_error</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row4_col0\" class=\"data row4 col0\" >0.02</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row5\" class=\"row_heading level0 row5\" >mean_squared_error</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row5_col0\" class=\"data row5 col0\" >0.00</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row6\" class=\"row_heading level0 row6\" >mean_squared_log_error</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row6_col0\" class=\"data row6 col0\" >nan</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row7\" class=\"row_heading level0 row7\" >median_absolute_error</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row7_col0\" class=\"data row7 col0\" >0.02</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row8\" class=\"row_heading level0 row8\" >mean_tweedie_deviance</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row8_col0\" class=\"data row8 col0\" >0.00</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294728_6b08_11eb_a2c4_e991cc078694level0_row9\" class=\"row_heading level0 row9\" >mean_poisson_deviance</th>\n                        <td id=\"T_16294728_6b08_11eb_a2c4_e991cc078694row9_col0\" class=\"data row9 col0\" >nan</td>\n            </tr>\n    </tbody></table>\n                                </div>\n                        </td>\n                </tr>\n                <tr class=\"no-border\">\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"2\" width=\"100.0%\">\n                                <img src=\"data:image/png;base64, iVBORw0KGgoAAAANSUhEUgAABJAAAAFxCAYAAAAs8FKpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFQAAAABJRU5ErkJggg==\n\" class=\"left big\" />\n                        </td>\n                </tr>\n                <tr class=\"no-border\">\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"2\" width=\"100.0%\">\n                                <div class=\"full-width\">\n                                    <div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead tr th {\n        text-align: left;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr>\n      <th></th>\n      <th>prediction</th>\n      <th>label</th>\n      <th colspan=\"10\" halign=\"left\">feature</th>\n    </tr>\n    <tr>\n      <th></th>\n      <th>Close</th>\n      <th>Close</th>\n      <th>(0, Close)</th>\n      <th>(1, Close)</th>\n      <th>(2, Close)</th>\n      <th>(3, Close)</th>\n      <th>(4, Close)</th>\n      <th>(5, Close)</th>\n      <th>(6, Close)</th>\n      <th>(7, Close)</th>\n      <th>(8, Close)</th>\n      <th>(9, Close)</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>4731</th>\n      <td>-0.020364</td>\n      <td>-0.036910</td>\n      <td>0.012831</td>\n      <td>0.006216</td>\n      <td>-0.006099</td>\n      <td>0.018227</td>\n      <td>0.016311</td>\n      <td>-0.027888</td>\n      <td>-0.024106</td>\n      <td>-0.000233</td>\n      <td>0.034835</td>\n      <td>0.010158</td>\n    </tr>\n    <tr>\n      <th>4732</th>\n      <td>-0.019138</td>\n      <td>0.009419</td>\n      <td>-0.036910</td>\n      <td>0.012831</td>\n      <td>0.006216</td>\n      <td>-0.006099</td>\n      <td>0.018227</td>\n      <td>0.016311</td>\n      <td>-0.027888</td>\n      <td>-0.024106</td>\n      <td>-0.000233</td>\n      <td>0.034835</td>\n    </tr>\n    <tr>\n      <th>4733</th>\n      <td>-0.024113</td>\n      <td>0.018822</td>\n      <td>0.009419</td>\n      <td>-0.036910</td>\n      <td>0.012831</td>\n      <td>0.006216</td>\n      <td>-0.006099</td>\n      <td>0.018227</td>\n      <td>0.016311</td>\n      <td>-0.027888</td>\n      <td>-0.024106</td>\n      <td>-0.000233</td>\n    </tr>\n    <tr>\n      <th>4734</th>\n      <td>-0.024659</td>\n      <td>-0.009474</td>\n      <td>0.018822</td>\n      <td>0.009419</td>\n      <td>-0.036910</td>\n      <td>0.012831</td>\n      <td>0.006216</td>\n      <td>-0.006099</td>\n      <td>0.018227</td>\n      <td>0.016311</td>\n      <td>-0.027888</td>\n      <td>-0.024106</td>\n    </tr>\n    <tr>\n      <th>4735</th>\n      <td>-0.019915</td>\n      <td>0.004942</td>\n      <td>-0.009474</td>\n      <td>0.018822</td>\n      <td>0.009419</td>\n      <td>-0.036910</td>\n      <td>0.012831</td>\n      <td>0.006216</td>\n      <td>-0.006099</td>\n      <td>0.018227</td>\n      <td>0.016311</td>\n      <td>-0.027888</td>\n    </tr>\n  </tbody>\n</table>\n</div>\n                                </div>\n                        </td>\n                </tr>\n        </table>\n</div></td>\n                <td width=\"50%\"><style>\n    .left {\n        float:left;\n    }\n    .right {\n        float:right;\n        text-align: right;\n    }\n    .full-width {\n        width: 100%;\n        overflow-x: scroll;\n    }\n    .no-float {\n        float: none;\n    }\n    .no-border {\n        border: none;\n    }\n    .big {\n        width: 100%;\n        max-width: 100em !important;\n        max-height: 100em;\n    }\n    .va-top {\n        vertical-align: top;\n    }\n</style>\n\n<div class=\"full-width\">\n        <table width=\"100%\" class=\"full-width no-boder\" cellspacing=\"0\" cellpadding=\"0\">\n                <tr class=\"no-border\">\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"1\" width=\"50.0%\">\n                                <img src=\"data:image/png;base64, iVBORw0KGgoAAAANSUhEUgAAAYgAAAFzCAYAAAAkFp78AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdeYxl513w+e9z9rvVrb2X6nbb7XIc\nuwlynHYCL8wIEiwP46FRICRGKLKUII8UM3kxQjhiMUnmjTCaKOIPh1dYRDPNMNh6x2hi8o7sLI54\nXxTApnGY17gT4qXt7q5eart1t3PP/swft2+596pqV9Wte+7vI0VOVZ2uOqXuut865zyL0lprhBBC\niMsY/T4BIYQQO5MEQgghxFVJIIQQQlyVBEIIIcRVSSCEEEJclQRCCCHEVVn9PoHtNDk5yc0339zv\n0xgIYZJhKAB10Xs1mQbXkt8rhBhkYZLx5kILgEK4zOLi4lWPG6pA3HzzzRw7dqzfpzEQ5mo+mQbn\nohhEF6IxM1bs45kJId6N1+ebPPDki9wBPP3Qh3jgf/zZax4rvwqKq5oou0RJSpRkaK2JkowoSZko\nu/0+NSHEDerFAbpxmJ2uXPd4CYS4Ks82mRkrYijwo3T1ysGzzX6fmhDiBmw0DjBkt5jExvQiIYQY\nbDcSB5ArCCGEyLUbjQNIIIQQIrfeTRxAAiGEELn0buMAEgghhMidzYgDSCCEECJXNisOIIEQQojc\n2Mw4gARCCCFyYbPjABIIIYQYeFsRB5BACCHEQNuqOIAEQgghBtZWxgEkEEIIMZC2Og4ggRBCiIGz\nHXEACYQQQgyU7YoDSCCEEGJgbGccQAIhhBADYbvjABIIIYTY8foRB5BACCHEjtavOIAEQgghdqx+\nxgEkEEIIsSP1Ow4ggRBCiB1nJ8QBJBBCCLGj7JQ4gARCCCF2jJ0UB5BACCHEjrDT4gASCCGE6Lud\nGAeQQAghRF/t1DgAWP0+AdF/QZyy1AoJkwzXMpgou3i22e/TEiL3dnIcQK4ghl4Qp8zVfDINRcck\n0zBX8wnitN+nJkSu7fQ4gARi6C21QhzLxLEMlFI4loFjmSy1wn6fmhC5NQhxgD4H4vnnn+f2229n\ndnaWxx9//IqPh2HIJz7xCWZnZ/nQhz7EW2+9dcnHT548Sblc5stf/vI2nXH+hEmGbapL3mebijDJ\n+nRGQuTboMQB+hiINE15+OGHee655zh+/DhPPfUUx48fv+SYr33ta4yNjfH666/zyCOP8Oijj17y\n8d/6rd/i53/+57fztHPHtQziVF/yvjjVuJZcXAqx2QYpDtDHQLz00kvMzs5y8OBBHMfhgQce4Nln\nn73kmGeffZYHH3wQgI997GO88MILaN19Mfv617/OLbfcwqFDh7b93PNkouwSJSlRkqG1JkoyoiRl\nouz2+9SEyJVBiwP0MRBzc3Ps379/9e19+/YxNzd3zWMsy6JarbK0tESr1eKP//iP+cM//MNtPec8\n8myTmbEihgI/SjEUzIwVZRSTEJtoEOMAAzrM9fOf/zyPPPII5XJ5zWOffPJJnnzySQAWFha2+tQG\nUi8SQojNN6hxgD4GYmZmhlOnTq2+ffr0aWZmZq56zL59+0iShHq9zsTEBC+++CLPPPMMv/M7v8PK\nygqGYeB5Hr/xG79xxdd56KGHeOihhwA4fPjw1n5TQghxkUGOA/QxEPfccw+vvfYaJ06cYGZmhqef\nfpq/+qu/uuSYI0eOcPToUX7yJ3+SZ555hg9/+MMopfi7v/u71WM+//nPUy6XrxoHsTEyYU6IzTPo\ncYA+BsKyLJ544gnuu+8+0jTlU5/6FIcOHeKxxx7j8OHDHDlyhE9/+tN88pOfZHZ2lvHxcZ5++ul+\nnW7u9SbMOZZJ0TGJU81czZfnEULcgDzEAUDp3rCgIXD48GGOHTvW79PYkXqzqZ2LhrdGSbb60FoI\nsT6DFofrvS7KYHcByIQ5ITbDoMVhLRIIAciEOSHerbzFASQQ4gKZMCfEjctjHEACIS6QCXNC3Ji8\nxgEGdKKc2BoyYU6IjclzHECuIIQQ4obkPQ4ggRBCiA0bhjiABEIIITZkWOIAEgghhFi3YYoDSCCE\nEGJdhi0OIIEQQog1DWMcQAIhhBDXNaxxAAmEEEJc0zDHASQQQghxVcMeB5BACCHEFSQOXbLUxhCQ\nneKEWD+JwzvkCiLnejvFZRqKjkmmu5sDBXHa71MTYseROFxKApFzS60QxzJxLAOlFI5l4FgmS62w\n36cmxI4icbiSBCLnZKc4IdYmcbg6CUTOyU5xQlyfxOHa5FUi52SnOCGuTeJwfRKInJOd4oS4OonD\n2mSY6xCQneKEuJTEYX3kCkIIMVQkDusngRBCDA2Jw8ZIIIQQQ0HisHHyDEKskiU5RF5JHG6MXEEI\nQJbkEPklcbhxEggByJIcIp8kDu+OBEIAsiSHyB+Jw7sngRCALMkh8kXisDnkp18Aly7J0YkSTi61\neWO+SRin8hxCDBSJw+aRQAjgndnWcZrx1mIbFNw8WcK2THlYLQaGxGFzyTBXscqzTVzL4NbpCs5l\nt5aWWqEs1yF2NInD5pMrCHEJeVgtBpHEYWtIIMQl5GG1GDQSh60jt5iG2NVmTk+UXeZqPtC9cohT\nTZSkcntJ7EgSh60lvxYOqWvNnAZk/wgxECQOW0+uIIbUxTOnARxLrb5/ZqwoVwxiR5M4bA+5ghhS\n8jBaDCqJw/aRQAwpeRgtBpHEYXvJq8GQunjmtNaaKMmIkpSJstvvUxPiqiQO208CMaR6M6d7D6Pj\nNEMBZ1Y6MnNa7DgSh/6QQAyxXiT2jhbQWmNbpuwFIXYciUP/SCAEZ2o+y37E2XqH842ATGvZC0Ls\nCBKH/pJhrkPkahPjAE4u+7i2gR+lnI9STtd8ZqfK2JYp25CKvpE49J9cQQyJa02MO7PSwTYVi80I\nUFRcC9Mw+MHZ7lLfsg2p6AeJw84ggRgS19pS9Hy9g2UagAYNUZJR9yPmVtq8cb6J1sg2pGJbSRx2\nDgnEkLjWxDilFEmm2TNaIE4zztQ7JGnGLZNlgjRlsRVecsUgk+nEVpI47CwSiCFxrYlx0xUXBWRZ\n90phd7XAeNljvORSLToYSrHiR5f8GZlMJ7aCxGHnkZ/0IXGtiXF7x4ocnCpzvt7hh2frNDohJcdE\nKTgwUSLTGe0gkcl0YktJHHYmGcU0JHpzHs7UfOZqIVprdlULBHFKO0x4794qrmPQDlMaQcx7qhWq\nBQeFohFE+FGKaxmysqvYdBKHnUsCMWQ03eW7e3s9HD9TZ7riMVKwuX13lXP1DgqFHyUUHAtDwaG9\noxIFsSUkDjub3GIaEkGc8uqZFc41gtV5DY5loIFWGAPdq4zd1QK2qah3YtkLQmwpicPOJ1cQQ6A3\nByKMMiqeRZrBuXqH3dUCRdvED98ZpeTZJpMVj+kRT/aEEFtG4jAYJBBDoDcHouRZpBrsC6OQau2I\nsmdT92NOLrfJMo1hKEq2yUH5gRVbROIwOCQQQyBMMoqOyWjR4Vw9AMA0YKkdUfMj2lGCH3cfQodJ\nxrKp8ByLvaMFub0kNpXEYbBIIIaAaxm0goR2lBCnGQutgEYnpt1JmKq6TFc8DENxphawd8zDMQyW\nWt2RTvIMQmwWicPgkYfUQ6DkWpxabhPGGa6liOKMZhDznt1lbGWw4nevJIpu93mEfeHhtSyrITaL\nxGEwSSCGQDtM2DdewrEMFpoRJddiquKRaCh4FoYyWG7GuKYiSjOSVOOYhiyrITaFxGFwyS2mIRAm\n3dFLIwWbKM0o2CYLzZB2lLB3tMD5KCDVGfUgptFJOLPsM1FxaYcxe0ZlJJO4cRKHwSZXEEPg4nWY\nHNMgSTVF1yROMxabIY1OxLlGwD++vsi5FZ/Rko1jGcytBNT9UJb3FjdE4jD4+hqI559/nttvv53Z\n2Vkef/zxKz4ehiGf+MQnmJ2d5UMf+hBvvfUWAN/+9rf5wAc+wPve9z4+8IEP8N3vfnebz3ywXLwO\n02jRph3FtMOEEdeiHSbU2hFTRZebJkpMVws0ggSlFDdNFEk08hxCbJjEIR/6Fog0TXn44Yd57rnn\nOH78OE899RTHjx+/5Jivfe1rjI2N8frrr/PII4/w6KOPAjA5Ock3vvENXnnlFY4ePconP/nJfnwL\nA6O3DpOhINOwd7RIwTFJNJypt2kECadXfNphjGsZ7KoUaAYxtXbE+Xp3xJMQ6yVxyI++BeKll15i\ndnaWgwcP4jgODzzwAM8+++wlxzz77LM8+OCDAHzsYx/jhRdeQGvN+9//fvbu3QvAoUOH6HQ6hKH8\nlns9vUgcnCqzd7RAO0h57XyDhZWIimfi2iZztQ7fP1XjB+dWOF/v7k2NgpofyW0msS4Sh3zpWyDm\n5ubYv3//6tv79u1jbm7umsdYlkW1WmVpaemSY/76r/+au+++G9e9+hLUTz75JIcPH+bw4cMsLCxs\n8ncxeHrLbpxvdtAaSgWLZpjhOQYZivl6QBCmGArOrHTwbJOpiie3mcSaJA75M9APqV999VUeffRR\n/uzP/uyaxzz00EMcO3aMY8eOMTU1tY1ntzP1lt1wTQOtNRXPJkpS3lps4RhgKGhFCSXXYne1wIhr\nUfEsGe4qrkvikE99C8TMzAynTp1affv06dPMzMxc85gkSajX60xMTKwe/9GPfpS/+Iu/4NZbb92+\nEx9wva1HR4suowUHjUZrTRBlKKXwXIvxsst7dle4eaIESskucuK6JA751bef+nvuuYfXXnuNEydO\nEEURTz/9NEeOHLnkmCNHjnD06FEAnnnmGT784Q+jlGJlZYX777+fxx9/nJ/6qZ/qx+kPrN6Q1wMT\nRQxLsdCMSNIUpcA0Dd43U6Xi2fzobJNmEKNAdpET1yRxyLe+BcKyLJ544gnuu+8+7rjjDj7+8Y9z\n6NAhHnvsMf7mb/4GgE9/+tMsLS0xOzvLV77yldWhsE888QSvv/46X/ziF7nrrru46667mJ+f79e3\nMlAmyi6NIGahGaAyTZykLLUjLFMRJxlhlGIqWG6HvL3kM1F2ZT0mcVUSh/xTWmu99mH5cPjwYY4d\nO9bv09hWQZyubhDkWgYl1+LEQpNXzzYIo4x2lLDcDrl5okzdj+gkGY5pMFGy2TNa5L+/fVriIK4g\ncciP670uyo3lHOuNWMo0FB2TTMO/vL1MzU+oFhz2jRcwFNSaMa+cXqETpyRphm1B2XMYKzkyeklc\nQeIwPCQQOdYbseRYBkopHMugHnSX/LZNxdtLPmdWAlKV8W/zDU7VfGxDUfEcFpoBIzJ6SVxG4jBc\nZLG+HOttFBTEKSt+RDOIeXuxhWOZWKbiR+dbZFlKFKV4lsWKH1PyIpRhULAN/untZfaPFnEtg4my\nK7eahpzEYfjIFUSO9TYKOlcP6MQpjU5C0bVoBjFvLrZJ0gTHNGknKQVbkWWauaXukhv1dshKO0Ip\nRSfq3qqS2dTDS+IwnCQQOTZRdrujlQA/TEgzjWkYuI7JmWUfDSRaU7ItbNPCNhWZ1tiGIsWg7NlU\nCjbtKJHNg4aYxGF4yS2mHPNsk9GSSxAlqyOZdlU9JlOH75+ocWK+jWsb6FRjmCZJlnFgosB42SVD\nAwrLUHTiFNtU+JFcQQwbicNwk0Dk3IhnUXYtFtshaQqmoTi30qFSsFgJI/woIc3A0RlKQ5RqOnFK\nJ0xZbER4tsG+sYLMph5CEgchgci5ibLLXM3HMgyiOGa5ndAOY0quRcEy8BwTQ3eHxFY8G8swObnc\nJopSRgo2by40WWwFWIbirpvG+/3tiG0icRAgzyByr7fM90jBouDaZBo828JQsGvE48BYiZnRIpZp\nUG9HHD+7wko7plSwGSk4uKaJZxrM1YJ+fytim0gcRI9cQQwBzzY5tHeUuZpPxbM4udjih2dTFpsR\nBTfDj2LSTIPSKK3xoxjXKnLzVJk008w3fM41QkqeyU8cnGS06PT7WxJbROIgLiaByJHLl9W4fO5C\nFKf88EyDf357mTjJGC86/OB8nWaQMFZ0cEyD0YJDtezQiVLemG91Nw0C9o66BFHG999e5v0HxiUS\nOSRxEJeTW0w5cbVlNXpzF4I45c35Jkt+jKngfTOjzIwXmW+F3WW+DU0QJWgURc+i0Y5ZboYsNDq0\nowTDUEyPFBgp2JRcmzcWmv3+dsUmkziIq5EriJy4eFkNAMdSq+8HaMcpJdci1jBWsql3IqIspexZ\ndCJFO07IdEiapRQsi1hlzK10mK443DJRohWmTI8YeLZBzY/69n2KzSdxENcigciJ3rIaF7t47kKW\naSxLUbRNWmHKW4s+pGAoRZxlJElGoDV+nFBxbQ5MlknRjHgO0yMeRcei7idkKZQ9+WeTFxIHcT1y\niyknehsBXaw3d8G1DAxD0Y4SPMvgjfNNVvyIThqz6EekGWRo6kFGK8go2iaebVBxTM41QhaaAX4Y\n40cx5xo+t07Ji0geSBzEWiQQOTFRdomSlCjJ0FoTJdnqTnATZRdLwcklH8symBpxSdOMuZWAlh8T\nRDF+AGkKBRsSrWgGCaZpYJlwcqnNmWaIHyTsHy/JA+ockDiI9ZBA5ERvvoOhwI9SDMXqTnCebVIt\nusyMeoRJRiOI0QrSBJIEWiFEQAJkGQRJTJRoojhld9VjulzAvPA1an4ki/YNOImDWC8JRE6sNcS1\n+4zCIk4yoijj3EqAqSAEYsC88L9mBM1OTM0PCBOwTZMwywiTjCDNcC1TVnYdYBIHsRESiBy43hDX\n3sdX2iGtMOEHZ1Y4udxmJQhIsm4UAFIuXEEAOtN04oROkvDGQovFVkiWpdim0Z0vISu7DiSJg9go\nGY6SA9ca4nqm5uPaJieX2oRxxvlmSK0d0wi7C/TFaTcIAL3H2zaQaEgzsE0DzzAwFFQ8m3YQU3RN\nWdl1AEkcxI2QK4gcCJPuFqIXS7OMk8vdqwrDUIRpimUqNJqVdkTB6r7wX76hqGOBaxsUbQvHMknp\n7iGRZnDbrhH8MJWVXQeMxEHcKPkpz4GrDXFdaIWUXAvHMnAtkyTReLaBZSpMU3WvGPSVn6uVQBBn\nZFpjmXQX9au47KoWGC12Nw/qjY4SO5/EQbwbEogcuNoQ13aQMFXpvoiPlRziLGO+3r0V5aru9qLJ\nNT5fO4Ikzah3EizbYnfVo1qwaQUJnm2sjo4SO5vEQbxbEogcuNoQ15smSpiGsfrxkaLNQrNDzY8J\n0oyia6/+5avLPl8KnKnHLLV8bDTtMOXmiRK7qh6H9o5KHAaAxEFsBnlInROebTJRdleHuqI1i62Q\nJM1odGJefruGHyVMli3mV2xWOi0MBeir3mkiAc7VEgqOz/tuGqPgmFcMnRU7k8RBbBa5gsiJy4e6\nphrm6x2aQcJbiy1sS5ECFc/BsEw6iSa6WhkuooFaO+B8PaDRiVlqhTL/YYeTOIjNJIHIiYuHuiql\naIcJYyWXNNPsHilw594qSao5Plen2QmJwitvLV0uAtphyvlGh0YQXzG/QuwsEgex2SQQOXH5UNco\nzfBsg3aUkGjNj843mK93rwbaYYxpdp81XI8BkMJiK+JfTq5wZqV7hSKT5HYeiYPYChKInLh8qKtj\nGgRxhs40Z1Y6LDUjbMvAtgz8KCFJ176C0IBpdtdrGi12l/teagU0gmuNfxL9IHEQW0UCkROXD3Ut\nuRbtMCbVGgNNdGFy2/SIi2kYRFz94fTFNGBb3UiMFV0yNIYy8KN4G74jsR4SB7GVJBA5cflQ16Jj\n8v4D4xhK4VgGBdug4FiEcbbukUgmkGkYLVgo1d1cKNOaoi2D33YCiYPYahKIHOlFYu9oAYDldkTR\nNXEsk93VAnGqKXo2aarX9Rev6e4RUS7YtMKU0aLFZNllpGBv6fch1iZxENtBApEzlw933TdW5ORy\nm1o7wtAZ7aA7Uc5d6wEEF5YAN8FAceeeCnuqRZRCltnoM4mD2C4SiJy5fLjrVMXj0MwY7ShmyY+p\ntUM82yBZ6wEE3X8c0+UChqGo+fElmxCJ/pA4iO0kN5NzJkwyDAXn6u9sHrRrxGHfWJEfzK1wdiUg\nTDTrecys6S7zbSioFmxmxopbffriOiQOYrvJFUTeaM3pWqe7zLeCc42AV07V+ZeTK6z4CaAx1vm3\nbpuQ6gzbNHhjviWT5PpI4iD6QQKRN6q7wFIYp8w3QnQG7ShBKd19nmCotSdAXJBlEKQpu6sFqkVL\nZlL3icRB9IsEIof2jRVphjGZ1jiWQdG1cG2TEc+h7Dnr/kuPNBRti5snSowUHBzLkO1Gt5nEQfST\nBCJnXMvANAyqBYcDEyWmRzzKrsmIZ2FZiiBOaEXr+1wVB+7YM8LJ5U535VfANlV3tVix5SQOot8k\nEDnTm1GtgDjJiNOMqRGP/eNldKJZaaVrrsHUMz3isataZGaswPlG96pBthvdHhIHsRPIT3rO9CbL\nTZRdGkFMkmXMTlX48f3V7vLe63z+ADBZcnEshQk0g5goyWS70W0gcRA7xZqB0Frzl3/5l3zxi18E\n4OTJk7z00ktbfmLixnm2ycGpMj9x6yR7qwUy3Xu/QaWw/t8JGlHKaNFlyY+wDUPmQWwDiYPYSdZ8\ntfjMZz7DP/zDP/DUU08BUKlUePjhh7f8xMTG9WZRv7nQHZIK3Rf0g1Nlio5NEKcben5Qawe8cqpG\nlGb89HumrojD5V9PRje9OxIHsdOsGYgXX3yRr371q3ieB8DY2BhRtM6nnGLbXL7ExuVDUut+RMU1\nCTbwVxfGCYt+jGOoK64a1vp6YmMkDmInWjMQtm2TpilKdW9eLywsYKx3ppXYNpcvsXHxkNQVP+LE\nYgvjwjLf65VkoHSGZRlXDG293tcTGyNxEDvVmq/0n/3sZ/noRz/K/Pw8v/d7v8dP//RP87u/+7vb\ncW5iAy7fUQ66Q1IbQcK/nKyx2A5ZWu/41gs6Icw3I5p+dMWtqWt9PRkCuzESB7GTrbkW06/92q/x\ngQ98gBdeeAGtNV//+te54447tuPcxAb0dpRzrHdetONUU2+HnFnpcOJckyU/pDvPen1ioOYHLLVj\n0Jf+qWt9PRkCu34SB7HTrRmIkydPUiwW+YVf+IVL3nfTTTdt6YmJjZkou6sPpm1TEaeaKElphSmt\nIOZsPUCnGkdBuM5CaKAdxBRd88ISHmt/PVnQb30kDmIQrBmI+++/H6UUWmuCIODEiRPcfvvtvPrq\nq9txfmKdevMfllohfpTiWgYzY0WOz9VxLIPEUMRpuu6rB+gGohNnkHXXdlrP15MhsGuTOIhBsWYg\nXnnllUvefvnll/nTP/3TLTshceN6L9qXvM8xOX+mg5Fp2lHWnSy3AaMFi7l6wKmazy1T5UsCcLWv\nJ65P4iAGyYZvGN999928+OKLW3EuYpMFcYptGqRaYQD2DewUOlUuYJsKP0pkhNK7JHEQg2bNK4iv\nfOUrq/8/yzJefvll9u7du6UnJTbHUitksuKSZCmNTkK2wQFGBmCbJuNFl2iDk+zEpSQOYhCtGYhm\ns/nOwZbF/fffzy//8i9v6UmJdy+IU14/3+RcvcNyK8KPE7INzmHLAMcxGCnaRDJC6YZJHMSgum4g\n0jSl2Wzy5S9/ebvOR2yC3iznWiem3olRpkGqM0wT1rXX6EU6Ucr5RodDM6OySN8NkDiIQXbNXwmT\nJME0Tb73ve9t5/mITdCb5aw1tKIUy+z+NbeCjX+uxUaAaSgmCo6MUNogiYMYdNe8gvjgBz/Iyy+/\nzF133cWRI0f4lV/5FUql0urHf+mXfmlbTlBsXJhkFJ3ui3mzE9FohwRxutGLBwDSTGNoRZBlBHEq\nkVgniYPIgzWfQQRBwMTEBN/97ndX50MopSQQO1hvlrPSmvl6SCdKsQwTg4yNPmZuBBFhmlF2LJZa\noQxrXQeJg8iLawZifn6er3zlK/zYj/3Yahh6lNrArjNi2/VmOcdaYypN+8Js6hsZgxRnYKFIs4xG\nkDCz6WebLxIHkSfXDESaprRarUvC0COB2FmCOGWpFRImGa5lMFF2V2dRm6ZBO47INBtah6nHNmB6\nzMOzLfzoRm5SDQ+Jg8ibawZiz549PPbYY9t5LuIG9EYsOZZJ0TGJU80PzjboxAntKMGPM6JYUymY\nxM2UjU51K7s2SZqRaU3lRmbaDQmJg8ijawbialcOYudZaoVkuvvfKM2IooS3ln3Gii63Tpf5b3Mr\n/H8nU4Lkxh5Sj5ZdGkFCEKVorZmr+UyUXXlYfRGJg8iraw5zfeGFF7b8iz///PPcfvvtzM7O8vjj\nj1/x8TAM+cQnPsHs7Cwf+tCHeOutt1Y/9kd/9EfMzs5y++23881vfnPLz3WnagQJS60AraFgm5xc\nCYiSlExrCo7NnXtHmRxxSRNu6BlEvdNdIvxH5xqcawT804kl/vGNRVZ82VUQJA4i364ZiPHx8S39\nwmma8vDDD/Pcc89x/PhxnnrqKY4fP37JMV/72tcYGxvj9ddf55FHHuHRRx8F4Pjx4zz99NO8+uqr\nPP/883zmM58hTYdzq0s/ijGUQYZmoRlydqVDJ0ppht3rhVsnS5Qd+4biAHCuFvPKqRorQUzFsym7\nNsvtkO+frA399qISB5F3fVs74aWXXmJ2dpaDBw/iOA4PPPAAzz777CXHPPvsszz44IMAfOxjH1vd\ntOjZZ5/lgQcewHVdbrnlFmZnZ3nppZf68W30XdG2qHciXjm1wplah1Rr2kFKO0wIkxTHNlEGpDdY\niAw4W+/w1mKbc/UOi+2QONEsNsOhXrxP4iCGQd8CMTc3x/79+1ff3rdvH3Nzc9c8xrIsqtUqS0tL\n6/qzw8K1DJbbMY5lYJqKqbJDeKEG51c6nFz0yTIout1RTBtl0r1KWW5FnFhs41oGmc5odGIaneEc\n1SRxEMNizYlyg+7JJ5/kySefBGBhYaHPZ7MFlMJQmokRD8tQpBoKtkWUZrxd8zkwUWR2V5m3Fhvc\nyFOIblRMCpbJQiMg3TNCohVjJQc/Tjb5m9n5JA5imPTtCmJmZoZTp06tvn369GlmZmaueUySJNTr\ndSYmJtb1Z3seeughjh07xrFjx5iamtqC76T/bhovkmpNkGQoBbO7ysxOlblz9wgTJRdLgVYbnwMB\nkAKeaRAmMVGakWaaom0wUVzISSMAACAASURBVHYoOsM17FXiIIZN3wJxzz338Nprr3HixAmiKOLp\np5/myJEjlxxz5MgRjh49CsAzzzzDhz/8YZRSHDlyhKeffpowDDlx4gSvvfYaH/zgB/vxbfSdaxmM\nlz1GCza7Kh5TZZcsgyjpLqvx9kKTKNGYmNzIy3kClDyLZqyZGStQdi32T5TYO1pkxMv9BegqiYMY\nRn37CbcsiyeeeIL77ruPNE351Kc+xaFDh3jsscc4fPgwR44c4dOf/jSf/OQnmZ2dZXx8nKeffhqA\nQ4cO8fGPf5w777wTy7L46le/imkO57j8ibJLUPMZL7m0gphGkKCAXVUP2zQ49uYyzTBm/3iBThyz\nssHnyp4C01BMjrjcOlHm0EwV0zCIknRolv+WOIhhpfQQzYg7fPgwx44d6/dpbLqrLbVxZqVD0TH5\nP//+BG8ttrEtg79/fZ759saeQ1RsuHNvlV98/wwl12a64nLTRGloJstJHETeXe91cXjuEeSYZ5tX\nrLLaW9F1suKxeGE46o0soeU5JiOezS1TFTzLoFp0hmZFV4mDGHayh2ROTZRdoiRl76iHZxkcP1un\nE258FJOFxjIVcZpR9uyh2XZU4iCEBCK3PNtkouziWia7qgXiJCW7gclypqEouiZJkhKl2VA8d5A4\nCNElgcixdpiwb6zIRNllxHMoeWrDf+FaGewe6S73XbCM3D93kDgI8Q4JRI6FSUYrjAmSjIJjYloW\nzgY/h22Z3DJdoeiaN/YQY4BIHIS4lAQip4I4peaHvH6+RTOIiFONrbrzGjbCD0LeWmhR8+NcP3+Q\nOAhxpfz+xA+x3iZCI56DZxuYStHsxBQcm43eINIK/u1sg6VWmNvnDxIHIa5OApFDS60QxzJxLIPR\nkkOzE1Mp2HTilI3OJ9SpJkg0Vc/O5fMHiYMQ1ybzIHIoTDIMBecbISXX5r17u7Of/XAeP4zx/fUP\nZ8o0eLaBK3EQYuhIIHLItQzO1Ds4loFtGpRdm9ldFWrtiErBJjldo7bOJTcyusF5e7HFm2MF9o4V\nc3ElIXEQYm1yiymHJsou7SBBZxqtNZahWGyGTFddTENTLa5vLJMCxkoOY0WbaslhyY+Yq/kDv5Oc\nxEGI9ZFA5JBnm9w0USLVmrofsdQOKbkGnTCh3s4I1/kCr4CZ0RJ7RwvYpoHW4FjmQO8kJ3EQYv3k\nFlNO7R0toLXm5LLPYrPDQivGDxMKjoGfrC8QGvCjhFRrkkzjWga2qfCjwbyCkDgIsTFyBZFTvaU2\nXj/XpBVnKK3pxBnnWwF+sL7PYQHL7YA007TDhIJtEqd6IOdDSByE2LjB+0kX69YOE6arHlXPIko1\nC62QKMnWPVnOMqHo2JRcmwMTRepBPJD7QEgchLgxcospxxqdGFPByUWfxXbIcjsgyVIMuluJrsW2\nYLzkkmQZSZKBhpkBG8UkcRDixskVRE51l9qIGCk6TFU8oiSj2clwTbXu3wqSpDtk1u+ktKOEXdWC\nxEGIISKByKmlVshUxcO1DA5MFNlV9ijYkKaK9W4hWHQVnm1SLFi8vdghjDa6klP/SByEePfkFlNO\nhUlGxbNwLIM4zTBtk4miR7UEjWCFKFr7c1RchyBOKbsmUxWXejAYgZA4CLE5JBA51dty1LNNyq7N\n+2aqmMAPzzVwTQVrXEcoIENTcK3uYn9BzL/Odec/TFfcHTujWuIgxOaRW0w51dtyNEoywiTFNhXv\n3TPC+/aPUCis7/cCP4jIkowfnGuw1A7ZVfFwLIMz9YA355s7bka1xEGIzSWByCnPNpkZK2IoyDKN\nzjSGUpRdB0ut/Zu/DYyVPfwk7S7Y51jsHS/iWCYl16IdpztqRrXEQYjNJ4HIsV4k3n9gnGrRYbEV\n8L1/m2elc/2ZcjYwWjSYqniMFh0OjBcpWiau1Q2LZSiyTBMmN7DJ9RaQOAixNeQZRM4FF37Tr/kR\n/zq3wnInwrNtTOJrzoWIgTjLKDoWPzZTpWCbpBc9skgyjWGoHTGjWuIgxNbp/0+42DK9neUy3f3/\nhmFQtC0sY+2hrrYJkyUb1zKI0oyCbRDFKVGS0g4TSheW8ugniYMQW0uuIHJsqRWS6e5/l9sRUZIS\npxkNP1o7EIaJUgatKGH/aIlbd5VZaUcopdhb9fo+ikniIMTWk0DkWCNIqPshrmUxXXEpuxZ+mKCU\nwkBf8xaTCd3Z15mmaNlUPIs79lS389SvS+IgxPaQW0w55kcxhjKwLYPpiofCYKRo49nGdX8zMIBW\nlJGmGYnOdtR2oxIHIbaPBCLHirZFpjVxmpFquHmyQMWzKLo2lYKBp67+5xTQCWMW2yHn6wFVb2dc\naEochNheEogcGynYTJZdDAWNIGa06HDPzZNMVzxumaowUrjyz/TmWAdRQifU7K0WcJ13AtF78P3m\nQmtbtx+VOAix/SQQOTZRdlEKSo5FyTFpdhJc22C0YFN2TEZKHt5lf6b38DrVYBsay1Scr3cI4vSS\nUVFFxyTTbEskJA5C9IcEIsd6u8rNNwMs06DgGpgKDkyUmR4pMFX2qJQUFw9WNeheRSgDPNtg2Q9X\nQ3Cm5uNYJo5loJTCsYwt36Na4iBE/0ggcq4dJuwfL3Hn3io/cXAKwzCoFExSpUnTDNc2UGY3Chbd\nK4gIcExFlMF8PURrjWOZzDdDbPPSBxe2qbZsRrXEQYj+kkDkXJhkqy/q1aLDj81UsS0TNBQcE0t1\nRzSZQEI3EDYQpppmEOJaBoutbhi01sTppTMotmqPaomDEP23M4aniK2jNadqPlp3lwA3DHj9XIMs\n0eyuepxaaoEBTgamAqW6/01TAEWSaZb8iDjV7KoWiJLu8wbbVMSpJkpSZsaKm3rKEgchdga5gsix\nIE7pJBlhnGIphR8l/PPbK2QoxisOhjKxLBPLAMOELAOlwTTAswzQijhNiaLuEht7RwurK8T6UYqh\nNn+PaomDEDuHXEHk2FIrZMSzsU3F20ttTi351P0Ix4RMA2jKrsPJKCC+cOdI0w2FUzBwbLM7qumy\nEGz2FUOPxEGInUUCkWNhkmEoqLUjJksenSihFUQsZqDjhLeXfU4tNjBMKABJCjqDKIMk0ewf9dg7\nWmS04Gz5uUochNh55BZTjrmWwUIzxDFNbMsgjDUogzBOeXvJp9YKSDQULBgr2RRchWl137YsRStK\n8aOU3VVPhrIKMYQkEDk2UXZphwlBknC+3uF8s8Pr55twYY/pZpCSZmAoiLVGa41OwTJhxHVwLZMz\n9e7EOBnKKsTwkUDkmGeb7BrxOF8PCKIMtCLKMk4tt+nEKWXXouSa+An4fkLBMvAcSA0ATbVoUXFt\n3phvbsn5SRyE2NkkEDnnWgaTFY+JEQfHVNTbMX6YkqXdq4IgSYli8BNohhlag2daGKbJSiem7Nnd\nsa96rR0kNkbiIMTOJ4HIO6XYN1ZgqRlRa0e0o5QoSTAMCNOMMOn+I7AUxAm4jsXeqgtZyslFH8cy\neO/uSjcSm0TiIMRgkEDknGsZJKmmHSUkmWa0YKMMRRBn2IbCNsCxoeTCZNVGKUXNj8lQjHg2882A\nWjsi3KQF+SQOQgwOCUTOTZRdTtV8mp2ITpwwPeJiYuIHEQ0/ougYWJaB1qDTjLKrUErhWoqpik3F\nNUk1nFxqs+JH7+pcJA5CDBaZB5EzQZxyZqXD+XoHpRTTFZckSRkrOByPU1zL5I49Ff45DmkGKa5j\nM+pYNDoRcZIRZ4qqZ3PrVJmxkodrKQ5MlkhTzRsLTT5wYOKGzkviIMTgkSuIHAnilDcWWpxZ8XEv\nLMt9ph6w1I6wbYOfODhJwbFItGZ3pciIZxOnGUXbpOiYOLbJzEiBXRUXZRp4tsHNU2Vcy8SzDVpB\nckPnJXEQYjBJIHJkqRXihwklx8axTRzLpORaeLbFmXpA0bXYU/WI4pRMa3ZXXSZLDs1OhAnsrrhU\ni93nEErDiGehNZyu+Zxc7kZnoyQOQgwuCUSOhElGpru7wPVYhmKibOPZJudWOgRxxoHJErurRRzL\nwjINkkzTSTS1dkwGHJgsUnBM5psBRctEa40fpbiW2tDucRIHIQabBCJHXMvAUIrkoj0bkkzjORa3\njBcZr3hYhsK1TCzToOSYrPgRfpTghzGOqTi11GaxFZGlGdMjBWKtsU2Du/aPMlkprHvJDYmDEINP\nHlLnyETZZaUTU2uHFLUFF5blHi/alB2LmyZLvJiktMKEgmvw9lJIkkHJs4nTlMwAzzJZbsXYI92H\n2e/ZXV39/L0ribVIHITIB7mCyBHPNrl1qsze0SJhkhIlGXurHgenK4wUbEzDoOo5NDoxWQatMEFn\nGQ0/Jkk0aaowyVhuh6RZyg/ONS+5pbSe3eMkDkLkh1xB5IxnmxycKnNwqnzJ+yfKLm/ONzlT9xkt\n27QWY5JMowDbUihAGRlxajJVddk/WaLeDjm17LNvrIBpGGvuHidxECJf5ApiSHi2iedYjLgWZcci\nzjJunSozUnQxDQPXNCCFVpRQcW0KlsHNk2Vcy+BcPVhz9ziJgxD5I1cQQ+bW6QozScZyO2ah2UFr\nTZAkdKIU0phqwaXiWdTaEbdNj7B/vIgfyZWDEMNIAjFEXMug7NlE7ZADk0WmKg4naz6OaeDYBjU/\nwTMVhlKcrYd85/g5llsRh2ZGrvk5JQ5C5JcEYohMlF2Cms94ySWKU/7rawuUHZt7DhbphAlnGwHt\nTsJiK2TfmEeawhsLDVzXZM9YkdHipVuPShyEyDd5BjEkgjhlqRUSpZpGEDFZ8fjgLZPcsbuCYRho\nBe/dPcLusQIV16ITpURphjIN5usBf/ejeeZq/uqoJomDEPknVxBDIIhT5mo+jmUyVrQpuxZRknLr\nVAnbMsm05pXTdRabAUGUYJpQ9hyKjkErTFjpRFQ8k0xzIRIZD/7v/wRIHITIMwnEEFhqhTgXFu8L\n4pTz9Q5n6wFJmlHxbEquRcFWLPsR862YEc8gzTTzzYjxkoNpKDIUjmXw1lKbf//0v2AoJXEQIufk\nFtMQCJMM2+yuo3RyscX5RkjRMVEo5lsBcyvdq4tDe0cYKZgkaYYfxbSDmNNLPpnWTFc8Tix246C1\nljgIMQQkEEPAtQziVFNrR6wECZ045XwjYNEPKVgmSQqNToxhKH58ZpTpiodjmcSZBgVhlLDQCvnM\n//UyAE/86vslDkIMgb4EYnl5mXvvvZfbbruNe++9l1qtdtXjjh49ym233cZtt93G0aNHAfB9n/vv\nv5/3vve9HDp0iM997nPbeeoDaaLsEiUptXbIQiMgTVMMpUDD6eU2nTjBMBRRlBHGKUXX5papMj95\ncIJbporU/IT/8J9/gNaaL3/sx/nAzeP9/paEENugL4F4/PHH+chHPsJrr73GRz7yER5//PErjlle\nXuYLX/gCL774Ii+99BJf+MIXVkPy27/92/zwhz/k+9//Pt/73vd47rnntvtbGCiebTIzViRMM7QC\n0zKZqngYSqEMRZJmuJZJ4cL+D0qBYxoYpkGUKv7zv54D4PNH7uTfzU5ecza1ECJf+hKIZ599lgcf\nfBCABx98kK9//etXHPPNb36Te++9l/HxccbGxrj33nt5/vnnKRaL/OzP/iwAjuNw9913c/r06W09\n/0Hk2SYHJsrcNlmiYJkoBaahiC/sITFSsCg6JvrCVhLTFZc41fwff/8WpmHwpY8e4n0zoxIHIYZI\nXwJx/vx59uzZA8Du3bs5f/78FcfMzc2xf//+1bf37dvH3NzcJcesrKzwjW98g4985CPX/FpPPvkk\nhw8f5vDhwywsLGzSdzCYRjyLPWMldo24REkGWlP2bFzb5FTNZ6kVsrtSIMk0r5yp8799899QSvHY\n/3QHd+yprrmSqxAiX7ZsmOvP/dzPce7cuSve/6UvfemSt5VSKKWuOG4tSZLwq7/6q3z2s5/l4MGD\n1zzuoYce4qGHHgLg8OHDG/46edKbSb1ntMj0iIdrG5xd6eDZJouNEKUUByaK3DJV5vf/n1dRCv7X\nX7yTO/ZUMVT3zwshhseWBeI73/nONT+2a9cuzp49y549ezh79izT09NXHDMzM8Pf/u3frr59+vRp\nfuZnfmb17YceeojbbruN3/zN39zM0849pRSnl9sstSJKnsVEyeG18y1qfoQy4HyjwzMvn8E0Ff/+\nI7NUCzaNIOLWqYrcXhJiyPTlnsGRI0dWRyUdPXqUX/zFX7zimPvuu49vfetb1Go1arUa3/rWt7jv\nvvsA+P3f/33q9Tp/8id/sq3nPUh6s6ffXGgxV/NZ8SPmaj62aXDLVJmxssNyK6TeSUjQzIwWsA3F\nX754mijJ+NS/O8B40eHARInpC1uNbmQ/aiHE4OtLID73uc/x7W9/m9tuu43vfOc7q0NVjx07xq//\n+q8DMD4+zh/8wR9wzz33cM899/DYY48xPj7O6dOn+dKXvsTx48e5++67ueuuu/jzP//zfnwbO1Yv\nDpmGomPiRyn/5UfznFz2OVvvcHKpTd2PeXupzUIzoOLYLLYj/vLF04Dmv5sdJ8s0fpRwark7Uc6x\nzHXvRy2EyAeltdZrH5YPhw8f5tixY/0+jS3Xi0NvaY1z9Q5nax0cu7uEBigmSjb/5UcLnF3xGS26\n/N8vz6GB/+HOXdgm7BsrcNvuKmmqmR5xOTBRwo/SK3aqE0IMtuu9LsqwlBzqLa0BUGtHOKZJpWgz\nt9KhE2U0OglvL/uUPZNWlPGf/rk7Ouz9+0dIkgylFIZhYJsGnm2w2IrWtR+1ECJf5Cc+h3pLawBE\naYZlKmxDsdyKSLMUz4bz9YCFVsw/vFkDBR+7e4aKa6EMuHWyTMmxOVfvEMYpSZoRJamMYhJiyEgg\ncqi3tEaUZNiGohOlNMOE2ekKpmnSSWDZj3jmn+dAwcffvxfHVFRsE1MpHNtg33gBA4O5lQ67Rrzr\n7kcthMgnCUQO9ZbWMBR4jkWaZZRsi5smChRsg4Yf8dffPwPAL/34bm6eLlPybN53YIybJksst2OC\nKGGyYnPLZJm7bhqTOAgxhGQ/iJzqRWKG7qimV8+sEMQZcQb/8b++iUJx36FpLMskTTSWAc0g5eBk\nmSBJMU2T8bLLRNGROAgxpOQKYgh4tsmhvaOEccZ/+H9/gKHgsx++lffvq6JJmWt0WGyGLLUCFlsh\nI56FrRRJkhGm+pKtRoUQw0MCMSRO13x+56//GwD/y8/MMll2GSk5vGdXlamyS4Yiy2Cy7NAMEk4s\nt6kUbMaK9kVbjUokhBgmcotpCLw+3+SBJ18EFP/pf/4QcZLxxkITyzQpjJgstyOmyi4Vz6Lo2Fhm\nyp7RAkmqUUrhWN0hs0utkJmxYn+/GSHEtpFA5Nw7cWB1m9Dukhsm9SCiE2Voral4Nq5lEqcZo0Wb\nEc8mSDKg+wxjuRXSCGKgO0pKnksIkX8SiBy7WhwASq5FM4jIMk0UJ9TaEfP1gDv2jYCCIM4I4gAN\nvHY+o+5HjBRsqoV3bjfJsFch8k+eQeTUteIA0A4T9o4VeXvBpxWmjJdt4izjX082KF64nfTKXB0D\nCKKETMNCM6LoWDiWIesyCTEkJBA5dL04ADQ6MScX24yXHSqejVIGBybLTFRdTq8EaDQ/PjOKRtGK\nUkquxd4xj86Fh9S2qQgv3H4SQuSX3GLKmbXiEMQpNT+i3omZLDtoFH49YbzksMtUWKZJteDgWQZB\nkjFZdtEaLFOtBkLWZRJiOMhPeY6sFQfojkSaqngUHYtGkNIIElqdlDcWWigUJcfCMQ2COMO1DMZK\nDlGa0olSHFMRJbIukxDDQgKRE+uJA3RXeq14FnfsHWGhEdBoR4wWDJI044fnGiRpSjtMqLVDSo6F\naxmMl1ySLMO1TAyFPKAWYkjILaYcWG8c4J2VXi3D4EO3TjDfCFlqhVRdm73jxdV9JKarBQwFfpRS\ndEzuPjAuURBiyEggBtxG4gDdOQxzNZ92lDBWdBgpOBRdk4myS8mx6MQpN02UiJJs9WpBCDGc5BbT\nANtoHOCdRfw826AVJBgKqgWHkmORpBrH7P6TkJFKQgi5ghhQNxKHnt7ifXM1H8cy0VrTiVI0mt3V\nAiAjlYQQcgUxkN5NHHou3jPCtUySLGO85OJahoxUEkIAcgUxcDYjDj2re0aMdedHLLVC/CjFtQwZ\nqSSEkEAMks2Mw+V6sRBCiB65xTQgtjIOQghxNRKIASBxEEL0gwRih5M4CCH6RZ5B7GA3GofeA+cw\n6a6nJBv8CCFuhARih3o3cejNbyg6JnGqmav5TJRd2mFy1WhIUIQQVyO3mHagd3NbaakV4lgmjmVc\n2E/aoBOn/N2/zfP2YptGJ6ITdSMSxOlqUDINhoIzKx3+8Y1F3pxvElxY3lsIMZwkEDvMu33mECYZ\ntqlW3w7ilLcWWqQaKgUbjWKpHaF1Nya9oGRac74RYBkGIwWbJT9ajYgQYjhJIHaQzXgg3VuttafW\njogzzUjBRimFbRo4lkErjAmTbDUotXaEY5rYloFtGmiNbC0qxJCTQOwQmzVaaaLsEiUpUZKhtaYd\nJViGoui880zBMhR+2J0x3QtKlGZYF648kqy7DpMs2CfEcJNA7ABbsXxGby8Hzza4ZaqMUhBfiEYn\nSlGqG5NeUBTdj8dpRpRkjBYdWbBPiCEno5j6bCvmOVy8bEYQF5ir+biWSSuIaQQJCrhzb3V1pNJE\n2WWhGfL6Yotq0ebWyRKGUkRJKstvCDHEJBB9tB2T4HqxWGqFmIZi+hpDXGfGiuyueiw0Q87WA24a\nlwX7hBh2Eog+2c4Z0tdbiO/iYbFgcNOEtbqbnMRBiOEmgeiDrYzDRie9hUl2yQNs6O4m50cyvFWI\nYSdPILfZVsehN+mt6JhkmjXnMlw+LBZkNzkhRJe8Cmyjrb6tdLVZ1GvNZbh8WKzsJieE6JFAbJPt\neOZw+SxqYM25DJcPizUU8nBaCAHIM4htsV0PpHu3ixzrnUis53aR7CYnhLgauYLYYts5WkluFwkh\nNpMEYgtt92Y/crtICLGZ5BbTFunXTnByu0gIsVnkCmILyDahQog8kEBsMomDECIvJBCbSOIghMgT\nCcQmkTgIIfJGHlJvgp0eh42uzySEECBXEO/aIMRho+szCSEESCDelZ0eB7ix9ZmEEAIkEDdsEOIA\nN7Y+kxBCgATihgxKHECW8xZC3Dh5ldigQYoDyPpMQogbJ6OYNmDQ4gCX7kntRymudfW9pmWkkxDi\nchKIdRrEOPSstT5Tb6STY5kUHZM41czVfFnoT4ghJ7eY1mGQ47AeMtJJCHE1Eog15D0OICOdhBBX\nJ4G4jmGIA8hIJyHE1ckrwDUMSxxARjoJIa5OAnEVwxQHkJ3ohBBXJ6OYLjNsceiRneiEEJfryxXE\n8vIy9957L7fddhv33nsvtVrtqscdPXqU2/7/9u42tqm6DQP4NYYD5nzICqhsaNquZaQd7dyoL3Es\nk8JqQpwBhjA/iJE5NCFGIkMSHWMRM3w0kohGnGKCHzQmbLIY3sbEpSrCaJDohlHcWGBzjtKOl41R\n6bifD2TNak9lz/o21uv3iXLunt3nTsNF++/OX6+HXq/Hrl27Ao4XFRUhKysrbH3FazgQESmJSUBs\n3boVVqsVp0+fhtVqxdatWwNq3G43qqqqcOzYMTQ3N6OqqsovSOrq6pCSkhK2nhgORET+YhIQ9fX1\nWLVqFQBg1apV2LNnT0DNwYMHsWjRIqhUKqSmpmLRokU4cOAAAKCvrw/vvvsuXn/99bD0w3AgIgoU\nkzWInp4ezJw5EwBw7733oqenJ6Cmq6sL9913n+/xrFmz0NXVBQCoqKjAK6+8guTkW39mXlNTg5qa\nGgCA0+kMOM5wICJSFrGAWLhwIf7666+Av3/zzTf9HickJCAhISGgLpiTJ0+ira0N27ZtQ0dHxy3r\ny8rKUFZWBgCYN2+e3zGGAxFRcBELiMbGxqDH7rnnHnR3d2PmzJno7u7G3XffHVCTnp6OpqYm3+PO\nzk4UFBTgxx9/hMPhgFqthtfrxfnz51FQUOBXOxIMByKifxeTNYiioiLft5J27dqFJ598MqDGZrOh\noaEBvb296O3tRUNDA2w2G1588UX8+eef6OjowPfff4/Zs2czHIiIIiAmAbFx40YcOnQIer0ejY2N\n2LhxIwDA4XCgtLQUAKBSqVBRUQGLxQKLxYJNmzZBpVKF/LMZDsEN3dW13dnHfauJCAkiIrcuGx/m\nZudgUvF/ATAc/mn4Lb/vSEzAlWteOK9cQ2pyEv4z5Q7uD0E0Ts2bNw8Oh0PxWFzdaqPd2QeA4aBk\n+C2/Pd4bcPd7MHHCBHi8g7gh4DsKojgUVwEBMByCGX7L797+v5GUmIgpSYn4e1C4PwRRnIqrj5im\nT58OtVqteMzpdGLGjBnRbWgMuT54AyI3v3b8t3cQExISIAASAFy+6IJq2nTcEPAW4MPE+2smGM4l\nuLE4m46ODly4cEHxWFwFxL/5t8/h4h1no4xzUca5BHe7zYb/HSQiIkUMCCIiUpS4efPmzbFuYqzI\nzc2NdQtjFmejjHNRxrkEdzvNhmsQRESkiB8xERGRorgKiLG6k12shTKXq1evYvHixZgzZw6MRqPv\ntim3uwMHDiAzMxM6nU5xQyuPx4MVK1ZAp9PhoYce8ruzcHV1NXQ6HTIzM3Hw4MEodh15o53LoUOH\nkJubi7lz5yI3NxeHDx+OcueRFcrrBQDOnj2LlJQUvPPOO1HqeIQkjpSXl0t1dbWIiFRXV8uGDRsC\nalwul2g0GnG5XOJ2u0Wj0Yjb7fYdr62tlZKSEjEajVHrO9JCmUt/f78cPnxYREQ8Ho/k5eXJvn37\notp/uHm9XtFqtdLW1iYej0dMJpO0trb61XzwwQeyZs0aERH54osv5KmnnhIRkdbWVjGZTHLt2jVp\nb28XrVYrXq836tcQCaHM5cSJE9LV1SUiIr/88oukpaVFt/kICmUuQ5YtWybFxcXy9ttvR63vkYir\ndxBjbSe7sSKUuSQnBBiVuAAABf1JREFUJ+Oxxx4DACQlJSEnJwednZ1R7T/cmpubodPpoNVqkZSU\nhJUrV6K+vt6vZvjMiouL8c0330BEUF9fj5UrV2LSpEnQaDTQ6XRobm6OxWWEXShzeeCBB5CWlgYA\nMBqNGBgYgMczPn4zP5S5AMCePXug0WhgNBqj3vutxFVARHMnu9tJqHMZcvHiRXz99dewWq2RbTjC\nRnKtw2smTpyIqVOnwuVyjei5t6tQ5jJcbW0tcnJyMGnSpMg3HQWhzKWvrw9vvfUWKisro9rzSMVk\ny9FIGis72Y01kZrLEK/Xi5KSErz00kvQarWj7pPGt9bWVrz66qtoaGiIdStjwubNm7Fu3TqkpKTE\nuhVF4y4gxvpOdrESqbkMKSsrg16vx8svvxzOtmMiPT0d586d8z3u7OxEenq6Ys2sWbPg9Xpx6dIl\nTJs2bUTPvV2FMpeh+iVLluCzzz5DRkZGVHuPpFDmcuzYMezevRsbNmzAxYsXMWHCBEyePBlr166N\n9mUoi/EaSFStX7/ebzG2vLw8oMblcolarRa32y1ut1vUarW4XC6/mjNnzoyrRepQ5/Laa6/J0qVL\nZXBwMKp9R8r169dFo9FIe3u7b9GxpaXFr+b999/3W3Rcvny5iIi0tLT4LVJrNJpxs0gdylx6e3vF\nZDJJbW1t1PuOtFDmMlxlZeWYW6SOq4C4cOGCLFiwQHQ6nVitVt8/cMePH5fVq1f76nbu3CkZGRmS\nkZEhn376acB5xltAhDKXc+fOCQCZM2eOmM1mMZvN8vHHH8fkOsJp7969otfrRavVypYtW0REpKKi\nQurr60VEZGBgQIqLiyUjI0MsFou0tbX5nrtlyxbRarUye/bs2/4bXf802rm88cYbkpyc7HuNmM1m\n6enpidl1hFsor5chYzEg+JvURESkKK6+xURERCPHgCAiIkUMCCIiUsSAICIiRQwIIiJSxIAgCiIx\nMRHZ2dnIysrC8uXLcfXq1VGf69lnn8Xu3bsBAKWlpTh16lTQ2qamJhw5cuT//hlqtTro5vNEo8GA\nIApiypQpOHnyJFpaWpCUlIQdO3b4Hfd6vaM67yeffAKDwRD0+GgDgijcGBBEIzB//nz88ccfaGpq\nwvz581FUVASDwYDBwUGUl5fDYrHAZDLho48+AgCICNauXYvMzEwsXLgQ58+f952roKAADocDwM19\nBHJycmA2m2G1WtHR0YEdO3Zg27ZtyM7OxnfffQen04lly5bBYrHAYrHghx9+AAC4XC4UFhbCaDSi\ntLQU/JUmCrdxdy8monDzer3Yv38/Hn/8cQDAiRMn0NLSAo1Gg5qaGkydOhXHjx+Hx+PBo48+isLC\nQvz000/47bffcOrUKfT09MBgMOC5557zO6/T6cTzzz8Pu90OjUYDt9sNlUqFF154ASkpKVi/fj0A\n4Omnn8a6deuQl5eHs2fPwmaz4ddff0VVVRXy8vKwadMm7N27Fzt37oz6bGh8Y0AQBTEwMIDs7GwA\nN99BrF69GkeOHMGDDz4IjUYDAGhoaMDPP//sW1+4dOkSTp8+DbvdjpKSEiQmJiItLQ0LFiwIOP/R\no0eRn5/vO5dKpVLso7Gx0W/N4vLly+jr64PdbkddXR0AYPHixUhNTQ3fxROBAUEU1NAaxD/deeed\nvj+LCLZv3w6bzeZXs2/fvrD1cePGDRw9ehSTJ08O2zmJRoJrEEQhsNls+PDDD3H9+nUAwO+//47+\n/n7k5+fjyy+/xODgILq7u/Htt98GPPfhhx+G3W7HmTNnANzcGxwA7rrrLly5csVXV1hYiO3bt/se\nD4VWfn4+Pv/8cwDA/v37g+4lTjRaDAiiEJSWlsJgMCAnJwdZWVlYs2YNvF4vlixZAr1eD4PBgGee\neQaPPPJIwHNnzJiBmpoaLF26FGazGStWrAAAPPHEE/jqq698i9TvvfceHA4HTCYTDAaD79tUlZWV\nsNvtMBqNqKurw/333x/Va6fxj3dzJSIiRXwHQUREihgQRESkiAFBRESKGBBERKSIAUFERIoYEERE\npIgBQUREihgQRESk6H8Xq/tzn7x+HQAAAABJRU5ErkJggg==\n\" class=\"left big\" />\n                        </td>\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"1\" width=\"50.0%\">\n                                <div class=\"full-width\">\n                                    <style  type=\"text/css\" >\n#T_16294729_6b08_11eb_a2c4_e991cc078694row0_col0{\n            background-color:  #a5002695;\n        }#T_16294729_6b08_11eb_a2c4_e991cc078694row1_col0{\n            background-color:  #fffebe95;\n        }#T_16294729_6b08_11eb_a2c4_e991cc078694row2_col0{\n            background-color:  #00000095;\n        }</style><table id=\"T_16294729_6b08_11eb_a2c4_e991cc078694\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >0</th>    </tr></thead><tbody>\n                <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row0\" class=\"row_heading level0 row0\" >r2_score</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row0_col0\" class=\"data row0 col0\" >-7.35</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row1\" class=\"row_heading level0 row1\" >explained_variance_score</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row1_col0\" class=\"data row1 col0\" >-0.01</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row2\" class=\"row_heading level0 row2\" >mean_gamma_deviance</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row2_col0\" class=\"data row2 col0\" >nan</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row3\" class=\"row_heading level0 row3\" >max_error</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row3_col0\" class=\"data row3 col0\" >0.07</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row4\" class=\"row_heading level0 row4\" >mean_absolute_error</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row4_col0\" class=\"data row4 col0\" >0.02</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row5\" class=\"row_heading level0 row5\" >mean_squared_error</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row5_col0\" class=\"data row5 col0\" >0.00</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row6\" class=\"row_heading level0 row6\" >mean_squared_log_error</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row6_col0\" class=\"data row6 col0\" >nan</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row7\" class=\"row_heading level0 row7\" >median_absolute_error</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row7_col0\" class=\"data row7 col0\" >0.02</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row8\" class=\"row_heading level0 row8\" >mean_tweedie_deviance</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row8_col0\" class=\"data row8 col0\" >0.00</td>\n            </tr>\n            <tr>\n                        <th id=\"T_16294729_6b08_11eb_a2c4_e991cc078694level0_row9\" class=\"row_heading level0 row9\" >mean_poisson_deviance</th>\n                        <td id=\"T_16294729_6b08_11eb_a2c4_e991cc078694row9_col0\" class=\"data row9 col0\" >nan</td>\n            </tr>\n    </tbody></table>\n                                </div>\n                        </td>\n                </tr>\n                <tr class=\"no-border\">\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"2\" width=\"100.0%\">\n                                <img src=\"data:image/png;base64, iVBORw0KGgoAAAANSUhEUgAABIAAAAFxCAYAAAALXtNBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n\" class=\"left big\" />\n                        </td>\n                </tr>\n                <tr class=\"no-border\">\n                        <td class=\"no-border va-top\" rowspan=\"1\" colspan=\"2\" width=\"100.0%\">\n                                <div class=\"full-width\">\n                                    <div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead tr th {\n        text-align: left;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr>\n      <th></th>\n      <th>prediction</th>\n      <th>label</th>\n      <th colspan=\"10\" halign=\"left\">feature</th>\n    </tr>\n    <tr>\n      <th></th>\n      <th>Close</th>\n      <th>Close</th>\n      <th>(0, Close)</th>\n      <th>(1, Close)</th>\n      <th>(2, Close)</th>\n      <th>(3, Close)</th>\n      <th>(4, Close)</th>\n      <th>(5, Close)</th>\n      <th>(6, Close)</th>\n      <th>(7, Close)</th>\n      <th>(8, Close)</th>\n      <th>(9, Close)</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>6757</th>\n      <td>-0.021821</td>\n      <td>-0.003709</td>\n      <td>0.004458</td>\n      <td>0.002266</td>\n      <td>0.007750</td>\n      <td>0.002224</td>\n      <td>-0.001609</td>\n      <td>-0.003719</td>\n      <td>-0.000288</td>\n      <td>0.000738</td>\n      <td>0.007236</td>\n      <td>0.001456</td>\n    </tr>\n    <tr>\n      <th>6758</th>\n      <td>-0.021734</td>\n      <td>-0.008495</td>\n      <td>-0.003709</td>\n      <td>0.004458</td>\n      <td>0.002266</td>\n      <td>0.007750</td>\n      <td>0.002224</td>\n      <td>-0.001609</td>\n      <td>-0.003719</td>\n      <td>-0.000288</td>\n      <td>0.000738</td>\n      <td>0.007236</td>\n    </tr>\n    <tr>\n      <th>6759</th>\n      <td>-0.021371</td>\n      <td>-0.006707</td>\n      <td>-0.008495</td>\n      <td>-0.003709</td>\n      <td>0.004458</td>\n      <td>0.002266</td>\n      <td>0.007750</td>\n      <td>0.002224</td>\n      <td>-0.001609</td>\n      <td>-0.003719</td>\n      <td>-0.000288</td>\n      <td>0.000738</td>\n    </tr>\n    <tr>\n      <th>6760</th>\n      <td>-0.022148</td>\n      <td>0.006170</td>\n      <td>-0.006707</td>\n      <td>-0.008495</td>\n      <td>-0.003709</td>\n      <td>0.004458</td>\n      <td>0.002266</td>\n      <td>0.007750</td>\n      <td>0.002224</td>\n      <td>-0.001609</td>\n      <td>-0.003719</td>\n      <td>-0.000288</td>\n    </tr>\n    <tr>\n      <th>6761</th>\n      <td>-0.022015</td>\n      <td>0.002023</td>\n      <td>0.006170</td>\n      <td>-0.006707</td>\n      <td>-0.008495</td>\n      <td>-0.003709</td>\n      <td>0.004458</td>\n      <td>0.002266</td>\n      <td>0.007750</td>\n      <td>0.002224</td>\n      <td>-0.001609</td>\n      <td>-0.003719</td>\n    </tr>\n  </tbody>\n</table>\n</div>\n                                </div>\n                        </td>\n                </tr>\n        </table>\n</div></td>\n            </tr>\n            <tr>\n                <td colspan=\"2\">\n                    <pandas_ml_utils_torch.pytorch_model.PytorchModel object at 0x7f5c20948f10>\n                </td>\n            </tr>\n        </tbody>\n    </table>\n</div>"
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_name = '/tmp/regression_with_regularization.model'\n",
    "\n",
    "# in order to restore from a gobal namespace you need to wrap everything in a function\n",
    "def train_model():\n",
    "    with df.model(model_name) as m:\n",
    "        from pandas_ml_common.utils.column_lagging_utils import lag_columns\n",
    "        from pandas_ml_utils import FeaturesAndLabels, RegressionSummary, FittingParameter\n",
    "        from pandas_ml_utils_torch import PytorchModel\n",
    "        from pandas_ml_utils_torch.merging_cross_folds import take_the_best\n",
    "        from torch import nn\n",
    "\n",
    "        # the network provider needs to be wrapped in a function as well because we can't pickle nested classes\n",
    "        def net_provider():\n",
    "            from pandas_ml_utils_torch import PytorchNN\n",
    "            from torch import nn\n",
    "\n",
    "            class Net(PytorchNN):\n",
    "\n",
    "                def __init__(self):\n",
    "                    super().__init__()\n",
    "                    self.net = nn.Sequential(\n",
    "                        nn.Linear(10, 4),\n",
    "                        nn.Tanh(),\n",
    "                        nn.Linear(4, 4),\n",
    "                        nn.Tanh(),\n",
    "                        nn.Linear(4, 1),\n",
    "                        nn.Tanh(),\n",
    "                    )\n",
    "\n",
    "                def L1(self):\n",
    "                    # path to the parameters which should be regularized\n",
    "                    # the path is constructed from self.named_parameters() and allows the use of wildcards\n",
    "                    return {'net/0/**/weight': 0.02}\n",
    "\n",
    "                def L2(self):\n",
    "                    return {\n",
    "                        'net/0/**/weight': 0.02,\n",
    "                        'net/2/**/weight': 0.05\n",
    "                    }\n",
    "\n",
    "                def forward_training(self, x):\n",
    "                    return self.net(x)\n",
    "\n",
    "            return Net()\n",
    "\n",
    "        def loss(params):\n",
    "            from torch.optim import SGD\n",
    "            return SGD(params, lr=0.01, momentum=0.0)\n",
    "\n",
    "        fit = m.fit(\n",
    "            PytorchModel(\n",
    "                net_provider,\n",
    "                FeaturesAndLabels(\n",
    "                    [lambda df: lag_columns(df[\"Close\"].pct_change(), range(10))],\n",
    "                    [lambda df: df[\"Close\"].pct_change().shift(-1)]),\n",
    "                nn.MSELoss,\n",
    "                loss,\n",
    "                merge_cross_folds=take_the_best,\n",
    "                summary_provider=RegressionSummary\n",
    "            ),\n",
    "            FittingParameter(epochs=100),\n",
    "            verbose=1\n",
    "        )\n",
    "\n",
    "    return fit\n",
    "\n",
    "fit = train_model()\n",
    "fit"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [],
   "source": [
    "import time\n",
    "time.sleep(0.1)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "outputs": [
    {
     "data": {
      "text/plain": "-0.02199546806514263"
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.model.predict(fit.model, tail=1)[\"prediction\", \"Close\"].item()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}