{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# How to Use Anomaly Detectors in Merlion\n", "\n", "This notebook will guide you through using all the key features of anomaly detectors in Merlion. Specifically, we will explain\n", "\n", "1. Initializing an anomaly detection model (including ensembles)\n", "1. Training the model\n", "1. Producing a series of anomaly scores with the model\n", "1. Quantitatively evaluating the model\n", "1. Visualizing the model's predictions\n", "1. Saving and loading a trained model\n", "1. Simulating the live deployment of a model using a `TSADEvaluator`\n", "\n", "We will be using a single example time series for this whole notebook. We load and visualize it now:" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "Time series /Users/abhatnagar/Desktop/Merlion/data/nab/realKnownCause/ec2_request_latency_system_failure.csv (index 2) has timestamp duplicates. Kept first values.\n", "Time series /Users/abhatnagar/Desktop/Merlion/data/nab/realKnownCause/machine_temperature_system_failure.csv (index 3) has timestamp duplicates. Kept first values.\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "import numpy as np\n", "\n", "from merlion.plot import plot_anoms\n", "from merlion.utils import TimeSeries\n", "from ts_datasets.anomaly import NAB\n", "\n", "np.random.seed(1234)\n", "\n", "# This is a time series with anomalies in both the train and test split.\n", "# time_series and metadata are both time-indexed pandas DataFrames.\n", "time_series, metadata = NAB(subset=\"realKnownCause\")[3]\n", "\n", "# Visualize the full time series\n", "fig = plt.figure(figsize=(10, 6))\n", "ax = fig.add_subplot(111)\n", "ax.plot(time_series)\n", "\n", "# Label the train/test split with a dashed line & plot anomalies\n", "ax.axvline(metadata[metadata.trainval].index[-1], ls=\"--\", lw=2, c=\"k\")\n", "plot_anoms(ax, TimeSeries.from_pd(metadata.anomaly))" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "from merlion.utils import TimeSeries\n", "\n", "# Get training split\n", "train = time_series[metadata.trainval]\n", "train_data = TimeSeries.from_pd(train)\n", "train_labels = TimeSeries.from_pd(metadata[metadata.trainval].anomaly)\n", "\n", "# Get testing split\n", "test = time_series[~metadata.trainval]\n", "test_data = TimeSeries.from_pd(test)\n", "test_labels = TimeSeries.from_pd(metadata[~metadata.trainval].anomaly)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Initialization\n", "\n", "In this notebook, we will use three different anomaly detection models:\n", "\n", "1. Isolation Forest (a classic anomaly detection model)\n", "2. WindStats (an in-house model that divides each week into windows of a specified size, and compares time series values to the historical values in the appropriate window)\n", "3. Prophet (Facebook's popular forecasting model, adapted for anomaly detection.\n", "\n", "Let's start by initializing each of them:" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "# Import models & configs\n", "from merlion.models.anomaly.isolation_forest import IsolationForest, IsolationForestConfig\n", "from merlion.models.anomaly.windstats import WindStats, WindStatsConfig\n", "from merlion.models.anomaly.forecast_based.prophet import ProphetDetector, ProphetDetectorConfig\n", "\n", "# Import a post-rule for thresholding\n", "from merlion.post_process.threshold import AggregateAlarms\n", "\n", "# Import a data processing transform\n", "from merlion.transform.moving_average import DifferenceTransform\n", "\n", "# All models are initialized using the syntax ModelClass(config), where config\n", "# is a model-specific configuration object. This is where you specify any\n", "# algorithm-specific hyperparameters, any data pre-processing transforms, and\n", "# the post-rule you want to use to post-process the anomaly scores (to reduce\n", "# noisiness when firing alerts). \n", "\n", "# We initialize isolation forest using the default config\n", "config1 = IsolationForestConfig()\n", "model1 = IsolationForest(config1)\n", "\n", "# We use a WindStats model that splits each week into windows of 60 minutes\n", "# each. Anomaly scores in Merlion correspond to z-scores. By default, we would\n", "# like to fire an alert for any 4-sigma event, so we specify a threshold rule\n", "# which achieves this.\n", "config2 = WindStatsConfig(wind_sz=60, threshold=AggregateAlarms(alm_threshold=4))\n", "model2 = WindStats(config2)\n", "\n", "# Prophet is a popular forecasting algorithm. Here, we specify that we would like\n", "# to pre-processes the input time series by applying a difference transform,\n", "# before running the model on it.\n", "config3 = ProphetDetectorConfig(transform=DifferenceTransform())\n", "model3 = ProphetDetector(config3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now that we have initialized the individual models, we will also combine them in an ensemble. We set this ensemble's detection threshold to fire alerts for 4-sigma events (the same as WindStats)." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "from merlion.models.ensemble.anomaly import DetectorEnsemble, DetectorEnsembleConfig\n", "\n", "ensemble_config = DetectorEnsembleConfig(threshold=AggregateAlarms(alm_threshold=4))\n", "ensemble = DetectorEnsemble(config=ensemble_config, models=[model1, model2, model3])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Training\n", "\n", "All anomaly detection models (and ensembles) share the same API for training. The `train()` method returns the model's predicted anomaly scores on the training data. Note that you may optionally specify configs that modify the protocol used to train the model's post-rule! You may optionally specify ground truth anomaly labels as well (if you have them), but they are not needed. We give examples of all these behaviors below." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Training IsolationForest...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "INFO:merlion.post_process.threshold:Threshold 3.2263 achieves F1=0.5000.\n", "INFO:merlion.models.forecast.prophet:Add seasonality 15\n", "INFO:numexpr.utils:Note: NumExpr detected 16 cores but \"NUMEXPR_MAX_THREADS\" not set, so enforcing safe limit of 8.\n", "INFO:numexpr.utils:NumExpr defaulting to 8 threads.\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "INFO:fbprophet:Disabling weekly seasonality. Run prophet with weekly_seasonality=True to override this.\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "Training WindStats...\n", "\n", "Training ProphetDetector...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "INFO:merlion.post_process.threshold:Threshold 3.2263 achieves F1=0.6667.\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "Training ensemble...\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "INFO:merlion.models.forecast.prophet:Add seasonality 15\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "INFO:fbprophet:Disabling weekly seasonality. Run prophet with weekly_seasonality=True to override this.\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Done!\n" ] } ], "source": [ "from merlion.evaluate.anomaly import TSADMetric\n", "\n", "# Train IsolationForest in the default way, using the ground truth anomaly labels\n", "# to set the post-rule's threshold\n", "print(f\"Training {type(model1).__name__}...\")\n", "train_scores_1 = model1.train(train_data=train_data, anomaly_labels=train_labels)\n", "\n", "# Train WindStats completely unsupervised (this retains our anomaly detection \n", "# default anomaly detection threshold of 4)\n", "print(f\"\\nTraining {type(model2).__name__}...\")\n", "train_scores_2 = model2.train(train_data=train_data, anomaly_labels=None)\n", "\n", "# Train Prophet with the ground truth anomaly labels, with a post-rule\n", "# trained to optimize Precision score\n", "print(f\"\\nTraining {type(model3).__name__}...\")\n", "post_rule_train_config_3 = dict(metric=TSADMetric.F1)\n", "train_scores_3 = model3.train(\n", " train_data=train_data, anomaly_labels=train_labels,\n", " post_rule_train_config=post_rule_train_config_3)\n", "\n", "# We consider an unsupervised ensemble, which combines the anomaly scores\n", "# returned by the models & sets a static anomaly detection threshold of 3.\n", "print(\"\\nTraining ensemble...\")\n", "ensemble_post_rule_train_config = dict(metric=None)\n", "train_scores_e = ensemble.train(\n", " train_data=train_data, anomaly_labels=train_labels,\n", " post_rule_train_config=ensemble_post_rule_train_config,\n", ")\n", "\n", "print(\"Done!\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Inference\n", "\n", "There are two ways to invoke an anomaly detection model: `model.get_anomaly_score()` returns the model's raw anomaly scores, while `model.get_anomaly_label()` returns the model's post-processed anomaly scores. The post-processing calibrates the anomaly scores to be interpretable as z-scores, and it also sparsifies them such that any nonzero values should be treated as an alert that a particular timestamp is anomalous." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "IsolationForest.get_anomaly_score() nonzero values (raw)\n", " anom_score\n", "2013-12-14 16:55:00 0.424103\n", "2013-12-14 17:00:00 0.418938\n", "2013-12-14 17:05:00 0.484891\n", "2013-12-14 17:10:00 0.500257\n", "2013-12-14 17:15:00 0.449213\n", "... ...\n", "2014-02-19 15:05:00 0.419456\n", "2014-02-19 15:10:00 0.415807\n", "2014-02-19 15:15:00 0.406724\n", "2014-02-19 15:20:00 0.427094\n", "2014-02-19 15:25:00 0.428348\n", "\n", "[19279 rows x 1 columns]\n", "\n", "IsolationForest.get_anomaly_label() nonzero values (post-processed)\n", " anom_score\n", "2013-12-16 16:00:00 3.251397\n", "2013-12-16 18:35:00 3.681691\n", "2013-12-27 19:25:00 3.914430\n", "2013-12-27 23:20:00 3.260543\n", "2013-12-28 04:15:00 3.738462\n", "2013-12-28 06:20:00 3.303482\n", "2014-01-02 10:00:00 3.233514\n", "2014-01-05 17:50:00 3.791805\n", "2014-01-12 09:25:00 3.535895\n", "2014-01-13 10:05:00 3.314500\n", "2014-01-16 12:50:00 3.850349\n", "2014-01-24 12:50:00 4.170855\n", "2014-01-27 17:45:00 3.537919\n", "2014-01-28 22:00:00 3.451974\n", "2014-01-30 23:40:00 3.550075\n", "2014-02-02 23:45:00 3.359105\n", "2014-02-03 11:55:00 4.175556\n", "2014-02-05 05:10:00 3.675433\n", "2014-02-09 11:55:00 4.005116\n", "2014-02-13 19:15:00 3.247573\n", "\n", "IsolationForest fires 20 alarms\n", "\n", "Raw scores at the locations where alarms were fired:\n", " anom_score\n", "2013-12-16 16:00:00 0.701491\n", "2013-12-16 18:35:00 0.772563\n", "2013-12-27 19:25:00 0.810997\n", "2013-12-27 23:20:00 0.702972\n", "2013-12-28 04:15:00 0.781997\n", "2013-12-28 06:20:00 0.709952\n", "2014-01-02 10:00:00 0.698602\n", "2014-01-05 17:50:00 0.790835\n", "2014-01-12 09:25:00 0.748293\n", "2014-01-13 10:05:00 0.711750\n", "2014-01-16 12:50:00 0.800493\n", "2014-01-24 12:50:00 0.852493\n", "2014-01-27 17:45:00 0.748630\n", "2014-01-28 22:00:00 0.734366\n", "2014-01-30 23:40:00 0.750652\n", "2014-02-02 23:45:00 0.719052\n", "2014-02-03 11:55:00 0.853260\n", "2014-02-05 05:10:00 0.771522\n", "2014-02-09 11:55:00 0.825713\n", "2014-02-13 19:15:00 0.700873\n", "Post-processed scores are interpretable as z-scores\n", "Raw scores are challenging to interpret\n" ] } ], "source": [ "# Here is a full example for the first model, IsolationForest\n", "scores_1 = model1.get_anomaly_score(test_data)\n", "scores_1_df = scores_1.to_pd()\n", "print(f\"{type(model1).__name__}.get_anomaly_score() nonzero values (raw)\")\n", "print(scores_1_df[scores_1_df.iloc[:, 0] != 0])\n", "print()\n", "\n", "labels_1 = model1.get_anomaly_label(test_data)\n", "labels_1_df = labels_1.to_pd()\n", "print(f\"{type(model1).__name__}.get_anomaly_label() nonzero values (post-processed)\")\n", "print(labels_1_df[labels_1_df.iloc[:, 0] != 0])\n", "print()\n", "\n", "print(f\"{type(model1).__name__} fires {(labels_1_df.values != 0).sum()} alarms\")\n", "print()\n", "\n", "print(\"Raw scores at the locations where alarms were fired:\")\n", "print(scores_1_df[labels_1_df.iloc[:, 0] != 0])\n", "print(\"Post-processed scores are interpretable as z-scores\")\n", "print(\"Raw scores are challenging to interpret\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The same API is shared for all models, including ensembles." ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "scores_2 = model2.get_anomaly_score(test_data)\n", "labels_2 = model2.get_anomaly_label(test_data)" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "scores_3 = model3.get_anomaly_score(test_data)\n", "labels_3 = model3.get_anomaly_label(test_data)" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "scores_e = ensemble.get_anomaly_score(test_data)\n", "labels_e = ensemble.get_anomaly_label(test_data)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Quantitative Evaluation\n", "\n", "It is fairly transparent to visualize a model's predicted anomaly scores and also quantitatively evaluate its anomaly labels. For evaluation, we use specialized definitions of precision, recall, and F1 as revised point-adjusted metrics (see the technical report for more details). We also consider the mean time to detect anomalies.\n", "\n", "In general, you may use the `TSADMetric` enum to compute evaluation metrics for a time series using the syntax\n", "```\n", "TSADMetric..value(ground_truth=ground_truth, predict=anomaly_labels)\n", "```\n", "where `` is the name of the evaluation metric (see the API docs for details and more options), `ground_truth` is a time series of ground truth anomaly labels, and `anomaly_labels` is the output of `model.get_anomaly_label()`." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "IsolationForest\n", "Precision: 0.1667\n", "Recall: 1.0000\n", "F1: 0.2857\n", "MTTD: 0 days 23:31:40\n", "\n", "WindStats\n", "Precision: 0.2500\n", "Recall: 0.6667\n", "F1: 0.3636\n", "MTTD: 1 days 10:27:30\n", "\n", "ProphetDetector\n", "Precision: 0.3000\n", "Recall: 1.0000\n", "F1: 0.4615\n", "MTTD: 1 days 00:05:00\n", "\n", "DetectorEnsemble\n", "Precision: 0.4000\n", "Recall: 0.6667\n", "F1: 0.5000\n", "MTTD: 1 days 10:27:30\n", "\n" ] } ], "source": [ "from merlion.evaluate.anomaly import TSADMetric\n", "\n", "for model, labels in [(model1, labels_1), (model2, labels_2), (model3, labels_3), (ensemble, labels_e)]:\n", " print(f\"{type(model).__name__}\")\n", " precision = TSADMetric.Precision.value(ground_truth=test_labels, predict=labels)\n", " recall = TSADMetric.Recall.value(ground_truth=test_labels, predict=labels)\n", " f1 = TSADMetric.F1.value(ground_truth=test_labels, predict=labels)\n", " mttd = TSADMetric.MeanTimeToDetect.value(ground_truth=test_labels, predict=labels)\n", " print(f\"Precision: {precision:.4f}\")\n", " print(f\"Recall: {recall:.4f}\")\n", " print(f\"F1: {f1:.4f}\")\n", " print(f\"MTTD: {mttd}\")\n", " print()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Since the individual models are trained to optimize F1 directly, they all have low precision, high recall, and a mean time to detect of around 1 day. However, by instead training the individual models to optimize precision, and training a model combination unit to optimize F1, we are able to greatly increase the precision and F1 score, at the cost of a lower recall and higher mean time to detect." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Visualization\n", "\n", "Let's now visualize the model predictions that led to these outcomes. The option `filter_scores=True` means that we want to plot the post-processed anomaly scores (i.e. returned by `model.get_anomaly_label()`). You may instead specify `filter_scores=False` to visualize the raw anomaly scores." ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "IsolationForest\n" ] }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAGoCAYAAABbtxOxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAADBnUlEQVR4nOydd3gU1frHv1uSkEAIEAJJ6L33qiKCtCiISq8KFlSuiqjYLvYCwlWxK170RxNE8AIiRQERAVEpUqVDKAkQSKElJFt+fyRnnJ2dmZ2Znd2Zyb6f5+FhMzvl7JRzvvOet9i8Xq8XBEEQBEEQBEEAAOxGN4AgCIIgCIIgzAQJZIIgCIIgCILgQQKZIAiCIAiCIHiQQCYIgiAIgiAIHiSQCYIgCIIgCIIHCWSCIAiCIAiC4EECmSAIzXTr1g3//e9/NW178uRJlCtXDm63W+dWEXrCv8bz589H7969dd1/KPaphNq1a2Pt2rVhPy5BENaABDJBRCjhFgjC49WsWRNXrlyBw+HQZd+xsbEoV64c9y8jIyPo/SpF64vClStXUK5cOdx2220haJX+jBw5Ej/++KPp90kQBBEsJJAJgigVfP/997hy5Qr3LzU1VdX2LpcrRC2TZsmSJYiJicFPP/2Es2fPhv34BEEQhDgkkAmCwJEjR3DLLbcgISEBlStXxtChQ7nvtmzZgg4dOiAhIQEdOnTAli1bRPdx9OhR3HrrrUhMTETlypUxcuRI5ObmAgBGjx6NkydP4o477kC5cuUwbdo0nDhxAjabjROmGRkZ6N+/PypVqoT69evjiy++4Pb9yiuvYMiQIbjnnnsQHx+PZs2aYdu2bQF/1/Xr1/HEE08gNTUVqampeOKJJ3D9+nUAwIYNG1C9enW8/fbbSE5OxtixY+HxeDB16lTUq1cPiYmJGDJkCLKzswEABQUFGDVqFBITE1GhQgV06NAB586dw7///W/8+uuvePTRR1GuXDk8+uijis/77Nmz8fDDD6Nly5aYN2+ez3e1a9fGf/7zH7Rs2RIJCQkYOnQoCgoKuO+/+OIL1K9fH5UqVUL//v19LOY2mw2ffPIJGjRogPj4eLz44os4evQobrzxRpQvXx5DhgxBYWEhACAnJwf9+vVDUlISKlasiH79+uH06dOi7f2///s/dOnShfv7wIED6NWrFypVqoRGjRph0aJF3HcrV65E06ZNER8fj2rVquE///mPon3abDZ89tlnaNCgASpUqIB//etfECv4mpGRgdjYWO76AMDOnTtRuXJlFBUVyd6PQsaMGYPJkydzf7N7g3+sgQMHIikpCXXq1MEHH3wguh+CIEoPJJAJgsCLL76I3r17IycnB6dPn8Zjjz0GAMjOzkbfvn3x+OOP4+LFi3jyySfRt29fXLx40W8fXq8Xzz//PDIyMvD333/j1KlTeOWVVwAAc+fORc2aNTkr7zPPPOO3/bBhw1C9enVkZGRg8eLFeOGFF7B+/Xru++XLl2PYsGHIzc1F//79FQnRN998E1u3bsVff/2FXbt24Y8//sAbb7zBfX/27FlkZ2cjPT0dM2fOxIcffoilS5fil19+QUZGBipWrIh//etfAIrFbF5eHk6dOoWLFy/is88+Q2xsLN58803cfPPN+Oijj3DlyhV89NFHAIB+/fph6tSpkm1LT0/Hhg0bMHLkSIwcORJz5szxW2fRokVYvXo1jh8/jt27d+P//u//AADr16/H888/j0WLFiEzMxO1atXCsGHDfLZds2YNtm/fjq1bt2LatGkYN24c5s2bh1OnTmHv3r1YsGABAMDj8WDs2LFIT0/HyZMnERsbq+jcXr16Fb169cKIESNw/vx5LFy4EOPHj8f+/fsBAPfffz8+//xzXL58GXv37sWtt94acJ+MFStW4M8//8Tu3buxaNEirFmzxm+d1NRU3HDDDViyZAm37Ouvv8agQYMQFRUlez+qwePx4I477kCrVq1w5swZrFu3DjNmzBBtE0EQpQcSyARBICoqCunp6cjIyECZMmU4i94PP/yABg0aYPTo0XA6nRg+fDgaN26M77//3m8f9evXR69evRATE4OkpCQ8+eST+OWXXxQd/9SpU9i8eTPefvttlClTBq1bt8YDDzzgIxq7dOmC22+/HQ6HA6NHj8auXbt89nHXXXehQoUKqFChAu666y4AxQFgL730EqpUqYKkpCS8/PLLmDt3LreN3W7Hq6++ipiYGMTGxuKzzz7Dm2++ierVqyMmJgavvPIKFi9eDJfLhaioKFy8eBFHjhyBw+FAu3btUL58ecnftGLFCjz33HOS38+dOxctW7ZE06ZNMWzYMOzbtw87d+70Wefxxx9HamoqKlWqhDvuuAN//fUX97vuu+8+tG3bFjExMZgyZQp+++03nDhxgtv2mWeeQfny5dGsWTM0b94cvXv3Rt26dZGQkIDbbruNO1ZiYiIGDhyIuLg4xMfH49///rei67ZixQrUrl0bY8eOhdPpRJs2bTBw4EB8++23AIrvqf379+PSpUuoWLEi2rZtG3CfjOeeew4VKlRAzZo10b17d+53CxkxYgQn9L1eLxYuXIgRI0YACO5+5PPnn38iKysLL730EqKjo1G3bl08+OCDWLhwoep9EQRhHUggEwSBadOmwev1omPHjmjWrBm+/PJLAMVTy7Vq1fJZt1atWjhz5ozfPs6dO4dhw4ahWrVqKF++PEaNGoULFy4oOn5GRgYqVaqE+Ph4yeMkJydzn+Pi4lBQUODjN7x06VLk5uYiNzcXS5cuFW1/rVq1fFwRkpKSUKZMGe7v9PR03H333ZzQbtKkCRwOB86dO4fRo0ejT58+GDZsGFJTU/HMM8+gqKhI0e8TY86cORg5ciQAoFq1arjlllswe/Zsn3WEv/nKlSuiv6tcuXJITEz0OV9Vq1blPsfGxvr9zfZ17do1PPTQQ6hVqxbKly+Prl27Ijc3N2B2kfT0dPz+++/cuapQoQLmz5/P+VIvWbIEK1euRK1atXDLLbfgt99+U3xupH63kIEDB+K3335DZmYmNm7cCLvdjptvvhlAcPej8HdmZGT4/M633noL586dU70vgiCsAwlkgiCQnJyML774AhkZGfj8888xfvx4HDlyBKmpqUhPT/dZ9+TJk6hWrZrfPl544QXYbDbs2bMHly5dwrx583x8R202m+TxU1NTkZ2djcuXLwc8jhqE7T958qRP8J6wTTVq1MCqVas4oZ2bm4uCggJUq1YNUVFRePnll7F//35s2bIFK1as4Czccr9NjC1btuDw4cOYMmUKkpOTkZycjN9//x1ff/21omBB4e+6evUqLl68qOl8vfPOOzh48CB+//13XLp0CRs3bgQAUb9fPjVq1MAtt9zic66uXLmCTz/9FADQoUMHLFu2DOfPn8ddd92FIUOGqG5bICpWrIjevXvjm2++wddff41hw4Zx1yLQ/cinbNmyuHbtGvc3P2CyRo0aqFOnjs/vvHz5MlauXKn77yEIwjyQQCYIAt9++y0XmFWxYkXYbDbY7XbcfvvtOHToECfcvvnmG+zfvx/9+vXz28fly5dRrlw5JCQk4MyZM5g+fbrP91WrVsWxY8dEj1+jRg3ceOONeP7551FQUIDdu3dj1qxZGDVqVFC/a/jw4XjjjTeQlZWFCxcu4LXXXpPd58MPP4x///vfnPjMysrCsmXLAAA///wz9uzZA7fbjfLlyyMqKgp2uz3gbxNj9uzZ6NWrF/bv34+//voLf/31F/bu3Yv8/HysWrVK0e/66quv8Ndff+H69et44YUX0KlTJ9SuXVtxGxiXL19GbGwsKlSogOzsbLz66quKtuvXrx8OHTqEuXPnoqioCEVFRfjzzz/x999/o7CwEPPnz0deXh6ioqJQvnx57lzpzYgRIzBnzhwsXryYc69gv0vufuTTunVrrFy5EtnZ2Th79ixmzJjBfdexY0fEx8fj7bffRn5+PtxuN/bu3Ys///wzJL+HIAhzQAKZIAj8+eef6NSpE8qVK4f+/fvj/fffR926dZGYmIgVK1bgnXfeQWJiIqZNm4YVK1agcuXKfvt4+eWXsWPHDiQkJKBv374YMGCAz/fPP/883njjDVSoUEE0o8GCBQtw4sQJpKam4u6778arr76Knj17BvW7Jk+ejPbt26Nly5Zo0aIF2rZt65OtQMiECRPQv39/9O7dG/Hx8ejcuTN+//13AMVWxUGDBqF8+fJo0qQJbrnlFowePZrbbvHixahYsSIef/xxAMBtt92Gt956y+8YBQUFWLRoER577DHOepycnIw6depg9OjRfm4WYvTs2ROvv/46Bg4ciJSUFBw9elSzT+wTTzyB/Px8VK5cGZ07d0ZaWpqi7eLj4/Hjjz9i4cKFSE1NRXJyMp599lkuS8jcuXNRu3ZtlC9fHp999hnmz5+vqX2B6N+/Pw4fPozk5GS0atWKWx7ofuQzevRotGrVCrVr10bv3r19srg4HA6sWLECf/31F+rUqYPKlSvjgQceQF5eXkh+D0EQ5sDmDTSPRhAEQRAEQRARBFmQCYIgCIIgCIIHCWSCIAiCIAiC4EECmSAIgiAIgiB4kEAmCIIgCIIgCB5OoxsQDHa73SfJP6EOr9erOn+r3uTn5wMoLlwgSoBiBbI4HNq3NQgzXBPCF7om5sPS16SU9mlhuyal9PyFArM/JwUFBfB4PEY3QxJLC+QyZcrg6NGjRjfDsmRlZSEpKcnQNrCiDVLXseiHH2BLSFC9X29eHqL69g2qbUZghmtC+ELXxHxY+ZqU1j4tXNektJ6/UGD256R+/fpGN0EWcrEgCIIgCIIgCB6WtiAT1mfkyJFGN4EgCIIgCMIHEsiEociVfyUIgiAIgjACEsgEQRAEQZget9uNKw0awF22rPqNPR7YLl7Uv1Emxuv14qIJfrPD4UB8fDwcFguSJIFMGMru3bsBAC1btjS4JQRBEISZuXz5MirVqIGKCQmqszN43W5NwX1WxuVywek0VuZ5vV7k5OQgJycHFSpUMLQtaiGBTBhKWloaACAjI8PglhAEQRBmxu12o0K5cqZOXUb4YrPZULFiRVy4cMHopqiGslgQBEEQBGEJSBxbD6teMxLIBEEQBEEQBMGDBDJBEARBEEQIqFSpktFNIDRCApkgCIIgCIIgeJBAJgiCIAiCUMC///1vfPrpp9zfr7/+OqZMmYI+ffqgU6dOaNu2LZYvX+633S+//IK77rqL+3vChAmYM2cOAGDHjh3o2bMnOnfujL59+yIzMzPkv8MK3HfffahSpQqaN2/OLZs0aRIaN26Mli1b4u6770Zubm7Ijk8CmSAIgiAIQgGDBw/GkiVLuL8XL16MUaNG4dtvv8Xvv/+OH3/8Ec8++yy8Xq+i/RUVFWHixIlYsGABtm7dijFjxuDll18OVfMtxZgxY7B69WqfZb169cLevXuxe/duNGzYEFOmTAnZ8SnNG2EowpufIAiCIJQQk5io+z6vX78u+33r1q1x/vx5ZGRk4MKFC6hYsSKSk5Px9NNPY9OmTbDb7cjIyMC5c+eQnJwc8HiHDh3Cvn37cPvttwMoTmWnZLtIoGvXrjhx4oTPst69e3OfO3fujMWLF4fs+CSQCUOhAiEEQRCEFq6rqBKnZ6GQgQMH4rvvvsO5c+cwaNAgLFiwABcuXMDWrVsRFRWFhg0boqCgwGcbp9MJj8fzT9tLhLjX60XTpk2xceNGXdpmJVwuF9q3b8/9PW7cOIwbN07x9l9++SWGDh0aiqYBIIFMEKWSH3/8EampqT6+WwRBEETwDB48GI888gguXLiAtWvXYvHixUhKSkJUVBQ2bNiA9PR0v21q1qyJAwcO4Pr168jPz8f69etx4403omHDhsjKysLWrVvRuXNnFBUV4fDhw2jatKkBvyy8OJ1ObNu2TdO2b775JpxOJ0aOHKlzq/6BBDJhKJMmTQIATJ8+3eCWlC7GjBmD5s2b48cffzS6KQRBEKWKpk2b4vLly6hWrRpSUlIwfPhwDBgwAG3btkW7du3QqFEjv21q1KiBgQMHok2bNqhduzZat24NAIiOjsbChQvx5JNPIi8vDy6XC4899lhECGSt/N///R9WrFiBdevWhbQICQlkwlDmz58PgARyKOBP5xEEQRD6sWPHDu5z5cqVJV0ksrOzuc9TpkwRDSpr1aoV1q1bp38jSyGrV6/GtGnT8MsvvyAuLi6kx6IsFgRRSgkUbEIQBEEQZmX48OG44YYbcPDgQVSvXh2zZs3Co48+isuXL6NXr15o3bo1Hn744ZAdnyzIBFFKOXr0qNFNIAiCIAhNLFiwwG/Z/fffH7bjkwWZIAiCIAiCIHiQQCYIgiAIgiAIHiSQCYIgCIIgCIIH+SAThtKiRQujm0AQBEEQBOEDWZAJQ1mzZg3WrFljdDMIIuJZvXq1T0oqgiDEWbZsGWJiYnDgwAFD21GpUiXF63o8HkycOBFt2rRB27ZtceONN+L48eMhbJ31IYFMEARB4L777sPnn39udDMIwvQsWrQIN910ExYtWmR0UxTz7bffIjMzE9u3b8eOHTuwaNEiVKhQIah9ulwufRpnUkggEwRBEACAwsJCo5tAEKbmypUr2LJlCz777DMfgfzLL7+gV69eGDZsGFq0aIF7770XXq8XALB+/Xp07NgRbdu2xbhx47gc9Q0bNsTkyZPRoUMH3HDDDdi5cyf69u2Lxo0bY+bMmdzx+vTpg06dOqFt27ZYvny5X5vuu+8+LFu2jPv73nvv9VsvMzMTycnJsNuLZV/16tVRsWJFAMUzuZ06dUL79u3Rp08fAMUFTgYNGoR27drh5ptvxp49ewAAr7/+OsaOHYtu3bph7NixyMrKwtChQ3HjjTfixhtvxJYtW3Q5z2aAfJAJQ0lNTQUAZGRkGNwSgiA2b95sdBMIQhGOF16Abe9e5Rt4vYBTXvJ4W7aE+513ZNf5/vvv0bt3bzRs2BCJiYnYsWMH2rZtCwD466+/sHPnTqSmpqJbt27YsmUL2rVrhwcffBCrVq1Cw4YNuZmaxx9/HEBxCeo///wTTz/9NB544AFs2LABBQUFnJguU6YMvv32W5QvXx4XLlzAzTffjDvuuMOnxPKYMWPwwQcf4M4770ReXh5+++03zJo1y6fdgwYNwq233orNmzeje/fuGDFiBFq3bo2srCyMHz8ea9euRZ06dTg3q9deew2tW7fG4sWL8fPPP+O+++7Dn3/+CQD4+++/8fPPPyM2Nhb33HMPHn/8cdx00004efIk+vXrh927dyu/LiaGLMhExHDs2DF88MEHKCoq8vvO6/Vi1KhRVJ6ZiGg6d+5sdBMIwtR88803GDx4MABg8ODB+Oabb7jv2rdvj+rVq8Nut6Nly5ZIT0/HoUOHULt2bTRs2BAAMGrUKGzatInbpl+/fgCA5s2bo2PHjoiPj0d8fDxiYmKQm5sLr9eLF198Ee3atcNtt92GjIwMnDt3zqdNXbt2xZEjR5CVlYVvvvkGd999N5yCl4Hq1atjz549eP3112G325GWlob169fj999/R5cuXVCnTh0A//g1b9myBSNGjAAAdO/eHdnZ2bh06RLX5tjYWADF1vEnnngCHTp0wMCBA3H58mVcuXJFn5NtMGRBJkLOsWPHUFRUhEaNGhnWBq/Xiy5dugAApk6d6mOxLiwsRO3atQEAe/fuRcuWLY1oYkjIycnhptEIIhCl3aeQKD2433pL1fpetxu2hISgjpmdnY0NGzZg3759sNlscLvdsNlsmDp1KgAgJiaGW9fhcCh6ntg2drsd0dHRKCgowKFDh2C32+FyubBgwQJcuHABW7duRVRUFBo2bIiCggK//YwaNQpff/01vv32W3zxxReSx0pLS0NaWhqqVKmC5cuXo2fPnqrPQ1xcHPfZ4/Hg119/RZkyZVTvx+yQBZkIOV26dEH37t01bbsnPR3XRDoDtaxYsULyOyaOAfhNS1mddevWGd0EwgKwgfz//u//jG0IQZiY7777DiNGjMDhw4dx6NAhHD16FLVr1/axCAtp2LAh0tPTceTIEQDA119/jZtvvllyfea3zP7Py8tDUlISoqKisGHDBqSnp4tuN3r0aHz00UcAgCZNmvh9v3PnTs4w5PF4sHfvXtSsWROdOnXCpk2buIwWzMXipptuwsKFCwEU+1cnJiaifPnyfvvt2bMnPv74Y+7vXbt2Sf42q0ECmTA1/d54Ax/973+atmWuFNeuXfPzrezbty8A+HU23377raZjmRXm50YQcly+fNnoJhCE6Vm0aBHuvPNOn2V33XWXj5uFkDJlymDmzJkYMWIE2rZtC7vdjnHjxkmuL3RPGD58OOfnPH/+fMmZ2KpVq6JRo0a45557RL8/f/48BgwYgDZt2qBdu3ZwOp0YP348kpKS8PHHH2Po0KFo3749Ro4cCQB48cUXsWPHDrRr1w6TJ0+WNB69++673HqtWrXiggtLAzYve02xIHFxcTh69KjRzbAsWVlZSEpKCvlxWCDeqVOn0LlzZ2zZsgVRUVE+30kF6aWmpuKhO+7Av0ePFv3+4MmTqFS+PJIE6Wq8eXmo9eCDsu2aMGEC3n//fb/lRgYM6nFN8vPzUa9ePe5vKwRAnjp1CklJSaacpgvXc6IWr9cLj8cDh8MR9L6+++47PProowCscb+Y9ZoooeiHHzRN9Xvz8hBV8mJvRsJxTS5evIgGVavCpuGe18PFQojb7dbl+eNz8uRJ5ObmomHDhqr6w2vXrqFt27b4/fffkVDyO10ul58vslEcOnQIiYmJPsvq16+Pq1evGtSiwJAFmZDkww8/xMGDB3Xb31dffYUzZ87glltu8ftOrkABP1pXSK+nn8ZjIiJXCWLiGIDlC5dI+Z+ZmU6dOmH69OlGN8NSvPDCC1z0fLDk5+frsh+CCBaPx2OZAhb79u3TfZ9MRJ4+fVrxNuvWrUOrVq0wfvx4ThwTwUMCmZBkypQp+PLLL4Pax/nz57nPL730EgDgxIkTnK/UtGnT8OKLL6J58+aS+3DY5W/TQp0Di8aOHavr/sJBYWEhNm7cCACWzcTx6aefGt0ES7Fjxw5kZWXpsi8KziPMwvLly3HTTTcZ3QzFiAXMBQMzCF27dk3xNj169MDhw4fJpU5nSCATssydOzeo7XNyckSXv1USgTxq1CguzY0U9gACeVdJ8EMks2rVKgwbNszoZhBhRG5mRS3MX1/Kf5EgwkVeXp7s92bxCmXt0DulmVl+n55Y9TeRQCZCilxlLiaeAz089gBCoMjtVt+wAJw6dUr3fYYS/jm0qgUZkH6hIoo5efIkUlNTVVmXpHjppZc4Ycz+Z7lNCcIo5O5th8OB3CtXTCG4Ll68CCD8PvtW69+9Xi9ycnJ099UOB+bw3iZKLT///LPkd3/99RfOnDnDdTRS6Gkp+/vvv0VT4Ah56KGHsHLlSt2OG2pY53Pw4EFcuHDB4Naogz/Y5ebmUt5mGVghj/r16wedr/u///0vatasiQceeIAb5MnVggCKxd+uXbtw6623yq6XmpqKgwcPIj4+Xvc27N69G/Xr1/fJuRsfH4/so0eRJROzIonHAxtvX1oC2AoLC5Gfn4+EhARkZ2dzYv7QoUOi6xcVFXEB6Yzc3FzExsb65Ezmc/36da4PF+7X7XYjMzMT1atXV9TeUAQRasHhcITkHgk1JJAJWYIVpyyBuhgsnUwgpFwsrqgMLHrttdeQkJCA1q1b46+//vL57rvvvsM333zDpesRfm922DlasmSJ5XLZ8q31f//9N1fRiZBHj6nd9957Dw888ACXwokEMgEA06dPx5w5c2Sto+zFdtmyZRg1apRux65atSoAIC0tDZUqVcJeXjlph8OBcocPB50F5I8//sBdd92FHTt2IDk5WfE+brjhBqSnp+PFF1/E0aNH8fXXXwOQz8K0efNmrk/btGkThgwZAgCSLxYbN27k3OWE+z1w4ABXTU8JWVlZfpkjCOWEzMXivvvuQ5UqVXyCr7Kzs9GrVy80aNAAvXr18plif/zxxzmryI4dO0LVLEIlgwYNCtuxpKbNLktMud06caKq/T/wwAMA4GcZnjVrFjp37sz5RVuRB0tS2i1YsMDvO7OLHv4MArtGRGCOHTsW9D7cAvcks98rRHhQkp2BFb6QK8KkBX5qM7nsRsHALL9qC0OxvPmvv/56QHekKVOmAPB9xvbv3899bteuneh2ci4UDz30kOK2EsETMoE8ZswYrF692mfZ1KlTuWjLHj16cNbFVatW4fDhwzh8+DBmzpyJRx55JFTNIlSip3tDIJi/stJO8azMeqyTmTt3Ln788UeuM2fs3LkTK1euxIYNG3DbbbcBkPe/zM3NNYXfWyDcbjdnTWZTeEIRZDbkkuwToYHdE8J7g/kiE5GNkhgMNv3PsucAxS+7wb5khfMl7cyZM5q3DRQ8/uGHHwKAj4vDK6+8wn2+cuUKjhw5gk2bNvnMBgmfybNnz+L69esAgMOHDwOwbtCb1QiZQO7atSsqVarks2zZsmW49957AQD33nsvli5dyi2/5557YLPZ0LlzZ+Tm5iIzMzNUTSNUoLdAlstvWVBQgBkzZqB58+a4ePEi1wmUiY5WfZyzubkAgBtvvBHNmzf38WMDiqfxWrdujYYNG/osl8rF27RpU1SrVk11O8JNbm4u13GPGTMGgPkF8pw5c4xuQsTBcn0Lk/STBZkAlPUZ8+bN81vWokULrkqpVpgYBEIXNPrmm28CAKdBtKA03zwTtWJ07doVQ4YM8RmH+Bbk1NRUtG3b1s/tzGqBelYlrD7I586dQ0pKCgAgOTkZ586dA1D8FlejRg1uverVq+PMmTPcunxmzpzJlTJ0uVy65QGNRAJlDEhJScHVq1eDOsfCa3jp0iWfZexFKCUlBUePHsXcuXORkpKC3r17c59j4+NxQcTfmL8f4ffZRUVISUnBlStXVPlq9ujRg9sv/3eLLQsFWrM4sPa1atUKu3btAgDUqFEDKSkpWLZsGXr27KlbG/VGeI/8+uuvaNy4sUGt8UfrNUlLS0Pz5s3xn//8R7e2iPWJWu7JCxcu+NzT7HNUVJQl+lQrZztxFRbCpqEwi7ewEM4wXZty5cohOjoa58+flzSSOBwOn3soJycHKSkpuHDhQlD30LVr17j9xsTE+O1Lj/OXnZ2tqU8Xe/6k9sHWLSgo4L6X2p6/j/z8fNH1fvzxR275uXPn/IL/xLDyc2IGDAvSs9lsmqyT48aN4+qYx8XFWbbcqFmQO3+ZmZlYsWKF5trqLpfLZyZgz549SExMxJYtW/zeiDMzM/3yITNRV618eVQWsSTw9y38/prDgczMTE33B9sv27agoMBvWSjRcgyHw4HTp08jJiaGa+uQIUMwefJknDlzxtTPiXC26MEHH8SBAwcMao04Wu8jrfeg3D6FaNk//9lMSkriPl++fNnU9wofq7RTSFF0NGwaLKPewkJEhek3X716FRcuXEClSpUkMz2cP3+eu2/Kly8P4J/7MzMzU3OWFf69ef/99/tdZz3OH/85UnMfSc1sV65c2UfPHD16lFu3du3a3DHkZsbLli2LuLg4OJ1O0fXYjCBQHH/QpUsXRW226nNiBsKaB7lq1ao+D1CVKlUAANWqVfPxeTp9+rQlprNLM3r4OO3evZv7fPz4cS6aNiYmRlXKFy1tcdrtsm/raqhbt64u+wklLIr8xIkT3LJy5coBAN59910jmqSZS5cuGd2EoAmFjyC/KmWwSFX/Ih9kAvjHb5a54ojBv1dycnK4jA5A8exJoIIfUvBz55upBLpc5ohq1apxM3cA8Oijj6rePyst7fF4uEweUrBgQSK0hFUg9+/fH7NnzwYAzJ49G3feeSe3fM6cOfB6vdi6dSsSEhJ0EzeENvTomJjleciQIX45H9X4fnk0iA2XxxN0/ke9o7PlyM7Oxu+//655+9wSn2vCHPADl/Ti8uXLAPTxy1yyZInP3506dQJgfn91IjzcddddAP7x1RUjgZdq7YsvvvCLJWjSpImmF0W+8OaL7lAhFxfDZ8OGDbLfs2BvwLca4MKFC1W1w+12o3Xr1rLrrlu3TtE+ieAImUAePnw4brjhBhw8eBDVq1fHrFmz8Nxzz+Gnn35CgwYNsHbtWjz33HMAgNtvvx1169ZF/fr18eCDD+KTTz4JVbMIhQiDd7SwfPlyAMCzzz7r952SYh0MLZ2s2+NRnQSewSKNxSKcmd+83rz11lu4++67NW//2Wef6dgaIlj0qHQnhM2+DRgwIOh9Ce/tunXromfPnmRBJgAAFSpUAOA7IyXkjz/+4D5/+umnouvwA+6UIpc7PxSkpaUpWu/pp59WvE/+eRML0hs7dqzfslWrVgEoFsj8VHdihKJ/IfwJmQ+yWD5WQPzNx2az4eOPPw5VUwgN6PkABkqHEwgtAtkVRAWh9u3bAxC3ordp0yYkpUW1ZA84ceIEkpOTA3amVmHjxo3o2rWr0c3QBanKWsHA8h6PGTMG8+fPD2pfQjcW1l8naCjAQJQ+9MpmcvXqVdP3T2xmRilt2rTBzp07Fa+/efNmFBYW+lQ4HTt2LL766iuf9Vh60wsXLgR0p9LjJZnP1atX8eOPPwZlpCmNhNXFgrAOzA+sbdu2Qe8r2BKTUi4WD91xh+Q2wViQ2TRzMC4P4eDGG2+UTEtnRerXr290E3Tj7bff1n2frLpjs2bNfJZ37NhRt2No9RslShfvvPOOLvuR8nWXg7lempUpU6aork5Xu3ZtzvACFPd1wrF17dq1AIpnMH/77Tc0bdpUcn96lG3evn07UlNTARSn2vzXv/4V9D5LGySQCVHYVKselgRhDmLG888/r2h7KQuyWyYXZNalSzh48KCi/QthFY5++eUX3YKtdu/ejfvuuw/vvPOO6D615rXkV6FTgtfrDSo5PhF6jh496nc/LFy4UNKXkQX3EITZ0DJ+7NmzJwQt0Y9mzZpxmbS0cPPNNwMojnHJyMiQrIi5du1aZGRk+M1Y9ujRQ5dxecuWLdxnimERhwQyIQp7APmZKOTQYikYPny46HJhByllQZYL3vtl717V7WHwXULE0u1oCWRKS0vD6tWr8c4774hmaFm8eLHqfQLAXonfOXr0aNHlmzdvRocOHTQdiwgPN998MyZPnuyz7MknnwzZ8bS+SBKEGN27d+c+8zNSKOHEiRO6lFBXyw8//KB4XYfDgf/+97/c3/zPYgjzFQtjcpS4oPCNSevWrcO3336rpKmyMD9zQFnlxEiEBDIhitqOrW7durIVg8SoXLmy37K9e/ciMTERixYt4pZJWVc9Hg+G3Xqr6HedGjbUpTgGf1qM8euvvwa9X73Yv3+/6PLOnTuLLrdKcEdpsYrWrl1b1frsxfT//u//ZNfTY4qVwQQNS7tJEDfccIOm7fbu3etTYe7s2bOqtg9VELQYFStW5D4/+OCDqrblP39iBTtq1qzJZYYZP368z3eB3BbFXCsee+wxn783b96suK1S8AtoBVNRsDRDApkQRc3UPbN+KrU2y8HKkzdv3hwA0FlQCpqPx+NBnZQUxApSyAFAkdut2Qc5EKEQmcGmpAu0P/aSEapzohWWOokNyKzqXGmxaqqtZCUslgOIz85MmDBBc5vESElJ8RuEichF68tSpUqVEBcXh2+++QaAuswPwD8viC1atAAg/aKvB8FUmWMzOn379hUVyK1atcKYMWMwfPhwVK9eXdW++/btiyeeeMJv+fPPP49hw4bh/vvvxzPPPKOp3XwopWNgSCATokg57J88edJPIL722msAgOjoaG6Z2jzKBw4cwJEjR7i/K1SogIyMDNRPTZW2IHu9cDocot+73W5FpTiVwq9aFAqBfN9992nelpUo5WcgYHk02TlgnaHRAll47n7++WcA/xQzqVWrFgDtPtlm4M8//+Q+qw16E3vJFCuVrqUKqRxut5ssyASH0ucvJSUFgwcP9lvO/GzVxkgwgXz33Xfj+++/15QmTitq4k2Ym9wPP/zA9WF8PB4P7HY7HA6HaiHqdrtFn+/HHnsM7777LqKiorhy3sGIXJYpIxRFjUoLJJAJUYRC5l//+hdSU1PRuXNnzrrLYOlr+L67/ByZSihfvrxoMJ/dZoPU4+v2eCQFsqukg9KLhx56iPv83nvv6bZfRjB+d+xl5IEHHuCWsehkZpFlHanelmq11K9fH/fccw/3N5uOZYWBWIYGJT7tXq9Xl3zdcly8eFG1yGVBNYGqYQWCXTOx3MRq7+3MzEzunhDD5XL5TDkTkUd2djYXwOt2u2WLdbF786mnnkK3bt0k11MrcNlzn5iYiJiYmJAK5DvuuIPLPQxAtnovs4QzX2DmPgH4Fgh59913MXPmTHhKClXNmzePq/eghNzcXG5bKQoKCvDxxx+jVatWGDlypOz+duzYgdTUVNHUpMx3Wkv8UKRAApkIiNvtxv/+9z/ub6kHir2JFhUVcQF406ZNk933yJEjZR9yu80mac3wMguyWFYIr1dXgcwPPFFaeUkNaisj8d/6meWeP1CxDpZZd5gVUs9zohWWzogP+w0scIQJezkWLFiABg0aACg+H0ePHtWvkSW0aNECQ4cOVbUNEw+sbVph95lYxLra6xgod3d2djYaN26sap9E6aJnz55cAO/KlStFA5QZfMtl3759AYgX+FBr4WQvg1FRUYiJicH+/ftFZ1D0oHnz5mjVqpXPMpbZYciQIZg1axa3/PvvvwcA1KtXz28/5cuX5z7/5z//wSuvvCJpBZaibNmyAIA+ffrAG2Ds4scnBKrY+cILLwCQD8IT+01EMcaPloTpEbPSsTd7vitFUlISAGDgwIHcskAWy+nTp8vm8rXZbJLZKjgLssj3RzIyggpkaCjwfdZ7SjtY2AvLDTfcgN9++w2Ar8VS2F62DsulawR8oRfIBYefVF+KNWvWcJ83bNjATesaDSs88NRTT6naThiYyCxP/NkcNmirvR/ZS6bb7eZeroSBSewlhaZcIxM1AXXsftq8eTOio6ORkZERsDyyEtgsRrNmzRBTElsiJ9SDoU2bNn7LBg0ahL///hubNm3iRDEfoXDt2LGj6BiXmZmparaOjZmnTp1CTk5OWF1LCGlIIBOijBs3jrPmiU3xMqsB35r8008/we12Y9u2bdwyYVEDtdhtNkAqzVuJQPZ6vX6D+serVgWsRiSHsKJQMAKZdXYzZsyQXIcFfSgdpF599VUAxZYWNi0q5zfIKqcxoWyECOIPdLNnz5Zdl71sycHP56w2Wj6UsBdKVvhEaeXFX375xedvZs1isxc//PADN50rvB8DXU92/a9fv86t279/fx+3HDag61EcKBhmzpyJDz/80NA2EPKwF1ixYLJgYC+VDRo04ARyKPIix8bGSt7nI0aMAOD7TLHPQtHrdrtFrb179uxRNcvDNx7MmTMH77//vuJtlWA2A49VIIFMiFKtWjUMHDgQcXFxolO8YunFPvnkE9SoUcNnWcuWLWWPs3v37oDZL+TyINtsNlkrs1b4LhXBws7VkCFDsGvXLtEqTKzjVZrmiCV293g8eOuttwDIR54zVxc27R+KYiEFBQXIz8/npvOEgp3vRvLJJ5/I7ktJqWZ2XvlTkmawfjKBzIImr169it27dwcM7hTLxcr3G+ZbvISDb6DfzV7Sjhw5ws0+uN1uH/9J/j3ocrmwdetWLnj05MmTsj7MevLGG29gypQpYTkWIQ2z5vbu3dvvOzbbobff+smTJ7nPbEbj0Ucf1fUYQLHRRyqIm/XBfPcQ9rwJBfKkSZMkhXC44j1YJiA5xIwnZi8BbgZIIIcQl8ulOr+iWWCBAgkJCbJVe4J18E9LS0NaWprk93a7XXrw93qLg/i8Xq6q3pgpU5CpMnJaDGEgYjCw1HVAcacplXUDUF7ileXh5O9LrsNjgvqWW24B8I9FUU+GDBmCevXqcQEs1atX9xnw+Lm1lbhQBIINoPyBzO12o1mzZrjnnntUp1fSC/aSyAbO7OxspKWl4dNPP5XdTslAxxBahAL5ejJ3lCNHjnDp3DyCcuz8Af3q1auYNm0aVq9ezf2GcGHlDCZWITU1NWAfwATk3r17sWzZMp/vmO+rVJVUPYgRSd+pBx6PBy6Xi7v3pVKm7dixg/vMzhVfDO/evRtdu3bl/hY+k2osyPxnz2azqUrjyCzejFdeeQXTp0/HrbwaAczti49VtUk4IYEcQvLz81VV6DETJ0+eRE5ODpxOp+zgG+p8tXYZ6zB/6dESi+j6nTvxpw5t4nduwrKio0aNUrUvr9fLpS9zOByiLxzsHCt1C2GBIUpLjhYVFWHUqFGc+Pjoo48UbacGflU/9ns6d+7MWU7louID0b9/f0ycONFnGUu953a7uePl5uYiJycHa9euNUxolS1bFv369eMGTDY4ibkqaUU4+Ab6rcxlhS9o5ARyQUEBtm7dKnk8PZByiwnHLMDu3bsjVoiz87ty5UrZ9cqVK8d9fuSRR3y+Y25DSgQy/z5SQ2xsrKbtAsGsx+z5ZDmXxUhNTfUJ/uU/B6zQFTufwvvWbrfjgw8+UBSsyw9U93q9srEiycnJkt+tWLECM2fOxHvvvYcDBw5wy/llpRmUBzkwJJBDCHsArXgjfvXVV/j222/hdDplAwZCPcjYIF9Smp3jaN502XWVVQADwS/zmZiYqLrjdrlcnPiQeuFQ63PGop6Foqt+/fpo1KiR6Dbr16/nOvFQVE7izybwXW3YcrmOPRDbtm3jig8w2H3JDzwLlS+ymAVGCuEAzKahlYrMhQsXBlxHrUBm3/Pdmdq1a+cjkPn7FN5X7Du153f37t2YN2+ez7KWLVuiS5cuaNu2rah7VTgEclpamp/Pd6TA+p9A5cuFRTr4sSVys35CBgwYILp8+fLlfsKbT6j8ZoXuFYHc6VieeUDcbULqfnU4HLDb7T4VZvkWZz7CMUXOHWvOnDmS3wmNOQwWe8LHirok3JBADiFsUFJq5TMjDoeDe5s9efKk35utXI7Yr776Kujjy7pYlJCSmIgyvCIlerhY8OGLCKfTqfp6unlV/ex2u+aOKT8/n7MEMAuyUMhs3LhRNHE9wwirGeuc3W63TxnVYKPT2W9/8sknuSIzcoNHMKgpfV1YWOgzAKu1HMvlY2WoDdJj3//000/cspiYGMnCMcJS80wYqHWpmjp1qt8U9oULF7i832oLCukJ/7rMnz/f7zeXVpT2X0Jh279/f+5z5cqV0a5dO9ntA/ngPvzwwz6uG3LnXw+XLKC4/zt37pxPUSu73S4bB8M/tpqZFFYohI9UCW/henIxIs2bN1cc+Mtg8QPr1q1DamoqNm7ciM8//9xvv4QvJJBDCBNCVn5TczqdXMflcDj8AsGkKu4BxTkdg8UG6cGfLXfa7XDxzvEhmZyPWmCd1wMPPIC77rpL9fXk+7tJWZCZT7Ec8+fPx6BBgwD8M2XOd2tQghH3IvN1ExajCDbl3L333gug2BLFOnth5S69fi9/ujkQRUVFPgMwy+ks93LCv4716tULOFgJBTLb94EDB0StT2yad9++fT7LpQSyUAj36NEDgP+z6PF40KZNG83uEmI5sRmvvPKK7tbkVq1accKOL3YmTZqEnTt3Ii0tzaeQTTCcOXPGFEGjfG6//faAzwQTUzVr1pRcR8nvkstUI7Y9P7+vkA8++CDg8ZRQvXp13HzzzVxMBkNO+PKtsmLriQVds3WFwlcqMJDfXwChMWSsXLkSo0ePBgAMGzbM73u1aSkjARLIIURvC/LOnTuDqrimBYfDwVl52KDMpoPkotrFckhqwVYShCf3vd1u9+lQyurkuyZMoP7aa6+hfv36ktfz+vXrouckNzeXy7gg5YOsZAqfL1q0+udp6XhPnz4ddMW6wsJCfPLJJz65qdlgo7XMtliHLjyPeglkNdO9Cxcu9JnSZFZ/OSuVsBiCMM2dUKALB2r2O2+99VbRdIJSYkdqKvfhhx/mPrMXEf5xGDNmzMC5c+fwyiuviO4nUJEH4QzU119/zX2eOXMmevXqJbu9WrKysriZC+E1PXPmDHbv3i0r2tXQoUMHzf63oYD5toYrQ4iwOMXGjRu5/pFfeAoA7r//fsl7CPin6lswXJWZ/ZASrkLErOKVKlUSteiKCeT27dsrOr6SmZU77rjD52+5vi43N9cnpaMYaowAkQIJZA3k5OQoEhp6C+S+fftygUmhpm7dupgwYQL27duHnJwcn+++++47v/X5AxuAgNNvSrHb7dJBeiw3pd3OZbEAAKdO6XWWLVvmF9ywefNm7reuWrXKJ9pYSgy8/fbb3GcmbITnVAn8qklac9VqsWh17NhRVblUMdauXctlaWDW8rFjxwJQ577AR2y6f9OmTT5/B/Ps8V1A1Jy3nTt3ilb1Y0E9Ygj3/+677/r8zfJeM6QsyIC46JUaPKXSdB05coTLeMJ3yxDuh1nHly9fLrqf7du3iy5nCP2TWUlfhlg6Sb0QnvNdu3bpfgw9AzODhV27L7/80u87NfcMoOx5EOYv5geNCWd6+CWfQ8VlGdGpNCWbkvXYCyXzQebTsWNH0W2EAvmmm24KeBxhFgq5WCElL2pS7h+RDAlkDTRr1sxPEPK5cuUK3G43N2jxU11ZhYYNG3LRvcKOTqyT6NatG44dO4bPPvtMVWL31atXc6mkxLBB3uppQ4lfr8fDddpzf/wRQHHp1GCoVKkSateu7bOM+boCxVaPb7/9lvtb6nfwg1sYwik+JfDFn9frxZAhQ1TvQ2rQO336tGwGjSVLlqg+Fh9+XmOhJePGG2/0+fvll18GIO+TqJRLly75BNmogQnkypUrq7JE165dW/TeY2V8xWB+46xSXtWqVZGRkYF169Zh3rx5XOl2hnDg5b8IiIkX/jM0cOBAzopYs2ZN7N2718/1AvAvXAL8k0qwsLAwYF5npRw/flw3H1M1CFMl/vnnn7ofwwyl3RlyL4usqA2f5ORk2b480KyKUPTx70u+a49QLIfKkvmTzAuQUguyknueGU2EFmQ545bQxeLNN98MeBx+Ea4zZ85oMrrwMdO9ahbojGhErqBDw4YNUaNGDW4dscFHK3LFIPSEZV7o3r27XwlRqc6kTJky6N+/v6RPlhgtW7aULSZit9kgZatgyx0lLhYugYgJhbX92Wef9Yt6LiwsxPnz57nzIuzwxZCzRCgpM+p2uzXlCZV62ejYsaOmUrG5ubmKxP60adO4qHih2BMKR2YdZ0I5GNq2bYtWrVpp2pYN4lFRUaoEcufOnUX976Wu+YIFC7i0e0ILU5MmTXzymTLYYMYsy3/88Qf3nZhA5oujJUuW+FjtK1WqpLjgA3MdqV27tqio4nP16lVFKa5uuummgAWFGMePH0e/fv0UrRsI9vww0RMKkW4m0SH27O/cuVNy/ZiYGMm+XIkFWalVVphmjf8yLpUBQwuT58+X/E7KF1+IkjScbBzYtWuXz37lfNuFYyqrYisHf98dOnSQfQG3Kvfddx+qVKniE5ORnZ2NXr16oUGDBujVq1fQLwZymOfptRhKEuezVDjCt8NgCFfJSJYjlfnM8kVGuCoEASU+yHIWZJuNc7HYI/DPVhL4phaHw+EnliZOnIjWrVtz54VfnUyMMmXK+PgQCwebv//+O2A7tBZo0Tv4o2nTpj6ZKfgILcW33norHnnkET8/beE0PPPTC4VFTw386llqzpvb7RZ9iZw8ebLo+nxfapa+LxCsH2BFWQL5iAvvWa3WX6XWvby8PDRo0MAnxZUe3HTTTT4FHIKBnRM2U6HV1UcOMwXpib3kjRkzJuB2Ql9iRqCxSM6YI1dKnC+Y9fKXzgrwEq90XFUy9rHxvlOnTj4vSHLHEPYXSowfSkW9lRkzZozfzOzUqVPRo0cPHD58GD169MDUqVNDdnwSyBpRkw9UKi+tFsIlkJkFedu2bdi/f79fqjO9mDRpEiZNmiT5vU3GgsxwOBxwezxwCUSMnuedwbJQ8AuksIAT1skJB9o777zT5++EhASfwYpZ91iFM7k0bUDxQPf6669j7ty53DK5fKJ8QpXmTUwIvPTSSz6uFQUFBXA6nbLuAcA/93go/U+VwFws1Kbm42ct4SP2UsPKhANAnTp1FJd/ZeeI/c+v1igmSoXtl8swIAffD14ONW5WWtBDeLJnIZTp3eQyM4QbsXtYifuRlCgMNBaJZUpgKB0/9TJytJPID6wWJcHK/PMlrJAnhZaZhnBpASPp2rWrT98GFMcGMT/ve++9NyQ5/RkkkDWiJvhCzzzI4XooWO7e3Nxc/PjjjyETyPPnz8d8makvu80mKep8gvTcbhwU+HprDWSTg+VBFpsuE6tWBBT72PKr7wkzWbDPzBrID2YRQ0w4yqVk4sPOpZYOWW4qS/g8JCcnw+l0+lTaun79uui9o+ZlU8x9Zfbs2X5FRPSABbYIs6QEgl8YJhD8ioZK3BEY7PrZ7XY0bdrUp0rh5s2b/WYhhOJI64uSMKiOkZCQ4PM3C95Tg9SsiJgY1jKtKnRdkhPIwfpXszbzXxCNJpgZmZUrV/pMcyt5QeFbRa2c6pRPw4YNA67Dfrfb7VZsQQ43rVq18slUYxQulwvt27fn/s2cOTPgNufOneP6u+TkZFl312AhgawRtQOmFlatWoXU1FTunxg5OTmSU2DBsHnzZp/f+Pvvv3OfhYO/2qTlarDJlJouWQEutxsuj0f3CnpiMHErds75Fl0+QsEkLDbCksKz6WuxYD++eGCWTf4A1LdvX0XtX79+PZxOpyb/7BMnTkh+JwxEZKKXP0B8/PHHXAfI90NWM3gWFBT4PQ+9evXCzTffrHgfSmH3/+nTpxX5hTO+//570RLs7AWIIRQZal48+Rbk/fv3o1evXj6luIUC0uVy+fiTVq9eXfGxlCBM18b3ieYjJ6zq1q0rulwsZSQ/XaBS6tSpI9oWsWsrloVEDeye1tvFJBguXbokulyJ2BV7oQ4k+Pj3czDZPPTKiMTgV/lUS5MmTQKuw/plj8fjc47MIpAzMjKwatUqvPTSS0Y3BU6nE9u2beP+SVUClMJms4X0vJJA1ogwu4EcWqfwhKmdAN+HzOv1olmzZujUqZOiwDC1SKWG4Yu9UKeGscvkQWbL9x4/jg+XLPFJ9RYqxHyQhe0BikVvamoq3G63TyU9wF8gr1ixAsA/lYyKior8BhT+9XW73Rg9ejRee+010WPLsXTpUrhcLk0WRLmOSOlbPJui5Psn8qOxA6G2MApDy5Q8Kz7icrlUT5WLvTQKX0qE11iNJZ0JFv414VvRhS+xHo/HJxZC70FFLoctHzUluxlsIOdfQ7XTqmL9I3sGxCzXwVqlgi2CEwqkAu74z5RcgDX//Ct5noQCmW0j56Yg9pyNHz9e1xc6rTFBH3zwgc+MWCC0CuRwBeJblapVq3JGoszMzJCeLxLIKmHiSM0DK2VZFKNp06bcW1Sg9HD8VHPBFnIQQ6oj4Q++YjXe9SRgoZCS/w+eOhUWgay01DSLdn7nnXf8fFKF1fSYtYrvf1qrVi2f/QkTx3s8nrAGSwLyHbxcNLwY999/P/f59ttvV7zdp59+GnAdlm+ZT7C+12qsu+3atcPIkSP9lgtfrIRFf9QEn4kJZDlcLpdPZLyw/DNDmHJPKUpFhxbLLxO3fCGr9oVHbMaH3RNiFuRgny3WLyr1KQ8HUgGgfDcfFlguxG63+51zNWne+C+DcgHuvXv39lu2Zs0aXQMoFy5cKJrCMBByhbGEPPzww2jQoIEmgdy1a1fVbZNi/Pjxuu3LLPTv35+LoZg9e7ZfjI+ekEBWCfMPVdNBq6kql5uby1kUxeBbo/n+gKFIJySVzi3sWSykLMi8z263G/VUdGBaUZrR4MeSXMwzZszwc7E4dOiQT1ECltZI7hryrdA2mw1ut9unw5XL4ymW5kypYORb0uQ6eGGlK/4Mi1iaMv5gXb58eUVtAeTPERNBVatW9fsuWIGspohETk6OqKAWtiGY2ARhkF4g3G63Txo/qfRZWi3LSvtDLT657MWC/0KhNg5C7PrLCeRgLexsn1qzzYQCqbbwxyfm5sKKxDCE/bDaNG98gaz2WQwUk6GG5557DtWqVZP095cq5HHixAlVL48vvfQSYmNjfe4jJWN0t27dRCthSiFVjyE6OhoHDx7EHXfcoajvCqWbZDAMHz4cN9xwAw4ePIjq1atj1qxZeO655/DTTz+hQYMGWLt2bdBFrOQggayS9evXA5CuQBXqtD4XLlzA9evXMXHiRJ8bX89AQIYa4RIq7AF8kFkHVOR2I1EQKBQKmAU5UM5Jvg+mWNDW999/j/T0dNl98AcSl8vFBYj88MMPfhZkubyZYtY9qftU6Ev8/PPPc5+/+OILn+/kApn4ovyzzz7z+z42NhYZGRmqO2ZhtUJ+9g6Hw4GMjAy/gDFAW2EWPufOnVPs63/s2DEukJLvViF8RoXnUw1aBDI7P+np6QG3k7o2/fv391vWqVMnxf2ekn7q0UcfFV3OF1kxMTHIz8/HiBEjFFUJkxPILAZADz777DOsXr1alc96uBC2Scy3l6UX4z/3gLihItA9xP+e72IhnEl57733ZPejpxvf448/Lvu9MGMCQ6tbhloL8q233qrK2MVeFDMyMvxcuOLj42Gz2fzK1vPp1q2b4mMZwYIFC5CZmYmioiKcPn0a999/PxITE7Fu3TocPnwYa9eulbxmekACWSWsk5HyQw1HucZp06b5Re2HwtVBauqbL0CUBodJ0aJFC79E8XwCBumVkNaxo4+LRbkQTW2yIL1AEeH86zFt2jS/vJ87d+4MeK/wBzQmcIBid4tFixZx6eUCISYYpaw4QgHEF8HM0p2amoqbb75ZcSYBLZWxatSoIVo8QuiDLLR0ScEPMlVKly5duDRTFy9e9Auyk4P52vKttsI+g1+FUS1qXSz494/cbEOg/bFpZn7WDDXWXCXxGFIFSPjbHj58GPXq1cOGDRsUFZMQc1cbMWIEAOCJJ57w+07NdDq/Ta+99hruu+8+v6BFI9ixY4fP8yIUyMKXTeCf+0roH6/FggwUu/JUrFgRRUVF3P0vfKHZvn27XwCl8Njhgv+MMmuy0ip7YoQ6SE/q2VMqss1UyMaM0NlRCXuApAQy64j//e9/c8v0zscr5ocZijRXUh2DzWbjLKTBBgeuWbMGa9askfxeSZDewK5d0aFxY7zDOwf/Hjw4qHZJIfQf1kKPHj18/PDEBCzgO6Dx/ZhZxa9ff/1V0fHEopWVDnBSfvCBovz5hVC0cO+99yqa0lTqr6c2OhooFvZMRDECnTc2KyAMRmzdurXkfaM0hzUftYMtXyCrQRgo7PV6kZubywXJAMVuE/y0cnLnSIlAlpq54r/UCd1oxMQeH75VWok1UIuY4b+s8QWmFrGtB/369cPAgQO5v9W4ewjX1WJBBopfPnJycnDgwAHu+j399NM+6+zfv18ygwngn30klPDvMSaQhZU/1aBWIKudgZYSyErv32nTpvlULiR8IYGsEjZFGEgk1a1blxM+N910k27Hl6qwE2zeTj4swMzhcEimA2NBikqmN4PBJpMHmX3POu9tvNRaVULkbuF0Ov0C5pTAHzyFnZrUvSQUyEzgKKm0x0fMIqe0I5ZL7cYXSUKkshq88847io7rcDgMr0K2evVqLpMFI5CLAHsZYeXImStOcnKy5HV+8cUXAagTZfw8yGJlrcXa9eWXXwZcTxil/+CDD3KfR40aBa/X6xcQnJWV5ZPLnJ2j5ORkbhm7luzZEQtiZEi5L/HdiIRZJgJN0/P54IMPuM9SfUuwPuuhLpSiBeY2sWTJEkyZMkXWMiocZ7RakBmTJ0+WfKmWMhAwunbtqntaQin4133QoEHIyMgIqlJbqC3I/Jcv1if873//w+LFi33WW7t2ragbXmpqalhmva0KCWSVMN/jQAI5JiYmoDuGEKn1mN8zIG2Z0zMRO2u30+nEzJkzMXfuXBw5ckS3/avBbrdLu1iULGdWZhfvHNwq47YRDDk5OdwLBJ9AJVv5U/zCgUnK+sW34rAgvebNm2tKlSVEKACqVavGfRbL4SvGkCFDJL+TsgL269dP0b7tdjtcLhcOHz4sKlbeeust3crQquHatWu4du0aXC4Xzp49i3Pnzvlcj59++gnAPy9BrVq1wu7du9G1a1fuGfV6vaIVzPjuGIHgu1i8++67ePfdd7X+JB/ee+890SwgQHEhE4/HE1AcMYHMDyJmlmMWvDp9+nSuRPmyZcvwr3/9C2PHjsXGjRuRmJiIjIwMv3zgRUVFkmXN2axaoGw+a9eu9XEjkkqtp0Ygi700COMLQlnMQCmsn7nhhhtw77334o477vBbx2azYfv27Vy5d/5yLRZkxrlz5yRdivr37y97T4ll0AgV/CqmeghatQJZTRo5oFjgslgBZkDp1KkT2rRp47Ne06ZNyZ1CA3TGVMJy1QYSpNHR0dw0m1LxKhXY0bhxY+6z1Ju0ktrtSmGDg9PpRIUKFdCjRw/VD65S5IqgAMXiVy59mw3ifsqh6gykzrNU9DODL3aV+myKuVhozQPMh1nbfv75Z07c8YNOV65cqWgqnP0msSwZUii9j2w2G2bNmoVbbrkFixYt8jtnY8aM4cqNhpMmTZqgfv36qFmzJtq2bYs2bdqgUaNGfgKJ71dfuXJlzjUnLy8P69evFz1nakQAP0ivYsWKfmV9tQqKihUrilYLa9WqFTebI3yhmzJlik9lSdbf8V+Sli5dipycHIwfP94v3ViHDh3w73//G2+++abPbAffB93lcsHlcgX0B23QoIFoQQxWVVMosBcuXCi6HzUCuV+/fn5ZEYT3g1CwBIsWn3ohPXv29Ftms9l8/Mv5y4OxIMudz0CuP4FSfZoZNQJ548aNsuW5A8GvdBioLYQySCCrhHX+r776ql/EPx9+QQmlGSb4gkiqQ6lXr57ocjWFSwLB2muGN06HTJlf1mUKO9B7RHJp6kW1atUQFxfn96LCctjKlc1myA3y/JLRQguy2ECyd+9e1UUJ6tevD4/Hg5EjR4om5p8+fbpsqkEhaqKIlfrB8u+9J598MiRZWvTk/vvv9/E7FQ5Gdrsd8+fPR5MmTbBp0ybRfWgVyGIE6yIgdjwmkPkFagD/3ODsM9/Xd+LEifjqq6/w+uuvc+JRyYDNz7RQWFgo++y8/fbbAMTdzaRyTPNLYjPXiypVqqi6FmpzgAeL1+vF3XffHfQzIRUMKXdcPsEKLhY3I5YGUngcPQRyPZ7LT7hQI5Dr168fVArVZ599VtYljlCP8QrIYvA7pQULFkiuxxfPSi3I/HyP7LMwQEDYUbAMAWorS8nBfqMZ3jjtCn2Q+RbkNx54IGTtYSJBWHHq559/RkZGBud7GmgfUvAtpXyBLCw2AhTnWK5UqZKiSkI2mw0PPvggxo0b5zNlyfYpvK/UCCz+tKQUcXFxqFy5suJ9muHlDIDiEtYbN26UDQTkXzuhX7MWAmWx0NPlih2P3TdCFwyW2UV4bOF96fV6ZV8ixGCzHRkZGQEtyO+//z4A7fEY7NmtUqWK7i8YQOBAQqWwtgUrkNVkZxBeKz0Ea+3atZGRkYFKlSqF3MWidnIyZkmkDxQeK1Q5gUPdp9ntdtkAVDOM51bDHKOQheDnt2UieO/evZybQK9evfD444+LDhiB4FvzWCcorCzk8Xi4Qbt169aaKlMFwiziBCi2IAeqkCeX6UL39pTMDAiTrwfK6sBHrqPivxAJBbLQuiCWBk2KM2fO4NVXX8Urr7ziE/jocDhQWFjod48KBcJrr70m63MM+PqcCvnzzz9V3atmuQebN2+uOEhHLhuLEsuQnhZk4fVs0aKFT0YDtRQUFEgKFTELMhPH/BRqHo/Hx+9VTVT/tm3bUFRUpMg9adasWQHXEQbRvfnmm5zvc6h8XvVK/cZmjPR8CWLVyOQyF2l1sbjvvvswceJEjB49GgA4dxSl2T0CBWorwe12wxGgT3nsscd8/N7VCkqxPou/jASq9TDHKGQhPv74Y+7z4cOHART7bDLKlCmDJk2a+AyISt70CwsLfSoasWhvoe+a2+1G1apV8f7772PlypVISkrC7bff7hMxHixOp1NXl41gkHWxKOmkw+mjxned4TN27FjF++Cn5BPeGyylkc1m83vJEooDtVOkDL4AsNvt6Nq1Kw4cOMAFUAH+GSqWLVuGRYsWye5XLi9yxYoVuZzCSpATlFrFs9iA/Pvvv8uWsbXb7WjdunXQViUlbVYjAgJZkIX7atOmjV/glRr2799fHDDL2+9bb70FoLi/+O6775CamoqCggLs3LmTC2Tll7Tet2+fTx+nZOaD8f333yvyQQbELch2u93H51Y4A8Tvw9nvXLBggY9BRAxhvyM3g6RXQRL2Mq5kXBEG9Ir5FwPAHXfcgeXLl0u+RGkpNc1ISkqC2+3m7p2CggI8+uijPs94rVq1JLcPV5De888/z/nx1q5dW/WYKuYDHOosFmow+vhWhASySp588km/ZfxAD+Yryg/aUPKmLxSkP/zwAwD/QYQFq/A7l5UrV0pGYy9btkz11J7b7VYsQtSU39SCPYAF2Wazhd2CLCZkxPJ4JiYmcvcG37eYT0ZGBjp16sRND7PrWrlyZT+rnFA0qi23y4f/G1iuY34HKsyyIJe2igXLsfYGI8SUIKzyJUWZMmXw4IMPooxM0Zi7777bx8o5ceJEzJkzh7P26VVWXcrvWGuhHbUC2ePxBG2VFwrkMWPGICMjw6d/WbhwIe677z6f7dizwX8BA4BPPvkE27dvlz0mc8tZs2YNioqKEBUVhbvuukt2G7Gc8B6PB2vXrpXchh9Ux37nU089hddff132WMKUj8wSy+C7HwVqt1qk/KrlkLLa3n777Wjfvr3kFL3Qiqumv3U6nVyhEKfTiVOnTvll+Xj11VclS0rrYUFWOzps2bJFdYGjWbNm+WVeMZNAJtRDAlklYmnW2CDKpqodDodfBLZaZs6cKbrc7XZLClix6m6PPPKI6iIiYtZKMU6dOhVUNTAlyLlY8C3IHq8XY9LS8KDCNGJa4Xdy/NRaYvfFsGHDuBkAqYHWZrOhTJkynBBg91K7du1kLchKRaIYQgsyvy18+P6i/KwW/IwXZcuWRbNmzQD8Mz3L/g4GsfubvSyOHz9e0T4GDBiAV199FaNGjeKWiQVR2mw2pKamYurUqfjmm2/w3HPPcb9drxzmwrykDL6FVa5cuBC1QXp6CGS2H6H1dfbs2dznF154wW8bqb4kPj5e0qLJ4J8TJpDT0tI48dKjRw9Z1xYt8AVZIGEmLAI1WFCgSO8iUcA/44mWjDaBXkikCCbN28mTJzF79mx4PB5OgPNnEoDivkMqNaQeFmSv1xtygVqtWjU/tzczCWSjj29FSCCr5M033/RbxqwGHo+HE8hM6EhNyWulqKgIRUVFopYtofWCobZTVFp1y+FwBP3QTZs2DdOmTZP8Xmi1EsLSvHm9Xvzf6tX4XWURjWDgZxThp+1ixVW6du2K+Ph4ZGRkoFevXpL7YQM/UCwIGjVqhKioKB+BfPr0aZw8eZK7xo8qCDiRwmazcZX8pDpwr9cr6XrAb9dDDz3E3StsmlSPjlgo5po3b84VH1G6f7YP/r08adIkv/XY/vgFJNiArKXMrFilS2HpXgZzk/njjz8kX4rFUCuQ1cwKScGEyqhRo3zOo1R6SkYwMx3838cCVQ8cOMBZrWfNmhUwvZUS+MVW+H1OoAwV/L69WrVqXNo9PoHK0quF5f8Wy0ATCLmKdXIE48Y2d+5cXLt2DR6PR1MApV4udEYIRDMJZEI9JJB1gA08TCA7nU5uUHC73bJTe0ph/lButxurV69WXGaYbaMGl8uF3NxcVdtoZdSoUT4WPiEOmTzIPmneSj7vVhEsFyyNGjXCmTNncPz4cR+rKcuBq+S82+12bNmyhes8o6Oj8fPPP/tlBpg6dSq2b9+OTz/9FBkZGUF1tna7nZsa5gsp4T5Z+/nuP3l5eT5+jf369eMEaPPmzbFixQofq6hWWFuYtdJut+Ouu+7S5As8YMAA1duwAVnMXSQjI8PnnxCpF1U+HTt29LmO1atXV2VBDuRiIRQUXq83aHcRJpD5L3RAYDcUvQIu2XH5Lm3R0dEBn4Xvvvsu4L75rmJaLJZr1qzhqopu2bLF5zt+ER49uHjxIoDAL2986zzbRviirjQ/c7CFQgDtQYW6uFgYlEfZTALZ6ONbERLIQXLzzTfjo48+AlDcAbhcLtjtdlWDkZKHlwWBMdGkJN8h8yMNZOERMmnSJNkywuEkUBYLrtR0CNIyBaJ8+fKw2Wx+xUPYNLNUCdWJEycCKPYHZNdeeI1YYQnGo48+ittvv12XdvMDAPnH4HegZcqU4QalsWPHchZroUWtRo0a3D6cTifatm2rSuhJwUQV29fu3btV74NNbwtzhwutWGIDB/83MaTEhNo81ACCLu8qZkHmFxkQCgoxH3a1MMuqWoGsR3Eb4B8LMntW+NPZM2bMkNyO3budO3eWXIdfxTDQrJUYiYmJ3Hlgz30wlnMl8AtIicF3i2IZcYTFQYTuMlIEWygE0C6Q9QrSM0IemkkgE+ohgRwk/PRely9fRl5eHhwOh6rOUUkeWVaMgQkbftSvlM8n6yDVdvZs+j0czJs3TzY9mGypabYOz4JsBhwOBzZt2iRZOnjSpEnIyMjg3G+Sk5PRpEkTn3UWL17sky83Pj6ey3ARLHyLjNA/nllb27RpgzfeeANA8YDP1meD3K+//ooPP/wQZcuW5VwD9EzNxgYTqdLqgThz5gwXLCas3icMbhIbuMSsc1JVLPmBtI899pjoOsJp92BSrgHiFuTJkycDKK4aJxQjSlOkifH8889jzpw5kgJZqvyz3hQVFSE6Opo7Nr8NrFKeHHKWfb7I1yKQxTJy6F09j8FeyAMV2GBWY+CfIEThfa2myE+wFmStIlfL9fA7dlBba4d/joxOXUkCXT0kkHVk6dKl2L17NzZt2sR1uJUqVfIpOyuGEl+yQYMG4cUXX4TL5UJycjIefPBB7jspgay1U1FSZlgvnnnmGdkpecVBegZYkOWoW7duwA7J4XDg2LFjOHv2rGj0+LFjx7jPSv3ClWC32zlhzBfINpuNm44+e/Ys5xrUr18/jBkzxmf9unXrciKP+dfq2QEH8rFVsr3UttOnT/f5WzglDhSfb2FhEzkL2KlTpwAU+52LMWHCBJ+/pcS2UsQEMhM7FStW9HseCgsLZYsIyPHYY4+hZ8+e3HMmFMh6ppiUg4l8ZgXlt6F+/fqYO3eu7PbC54ef11sYrCo3Q5eamup3z4i9fGjNUBIIZqFOSkpSvA172RW+cD733HNYtWpVwO2DKRTCXEyEgXlK0cvFwmiBaPTxCfWQQFZBoGwUbBorOzub6zBjYmLgdruRn5+PX375RXQ79vCvXr2aG2iFxMTE4Oabb4bL5UKZMmV8OuRXX31VdBuWckbtm3s4BXIgAhYKYWnewtck3bDb7Vy6I6F4eeihh3wCofQIsuLDRK1QIDNeeukl7nN0dDRSUlJgt9tFywTzhbxesP0rqUyohCNHjnCfhUFTYs91UVGR3/mWe/4dDgfS09Mls148+OCDPr6wcqnnlCD1ApGRkYH4+Hjk5ub6pQkMdsqfTbMLrdFPPvkkF5iqBL4riBrYcZmlXPh7evToIbu98FyJZdwAIOoix84lu/8PHz4s67qWkZEhW1kxGFiOaanAT8A3bZ3X6+VedoUBjcnJyT4BxlIE44Mslv+Z/zwGglws9MHo41sREsgqECZdF8IsGh07duQ6b5aNoF69ehg+fLhoxDWzJrVs2RIOhwMdOnQQ3T+rWCUc7KR8XV9++WUAwC233BLgl/kiZQUzArlS01ynaYAFWQ//QofDwd1TwqlPYXUyvS3IDGGnKXZ/Op1ObppTLKf20KFDdWmXWBuZn2mwLwdCN4tATJ06VTJoUQq5oCmbzSbrA6sWOQu73W7HCy+84OOiI7T6aoHvYsF/oStTpgwWLVrkUylRLg1hICErBSsUwo4t9gyyvlPMAqz0HmJltRl9+/ZFzZo1UVhYyOXHXrVqlSaXJ75wDRZ+/m4hfPHOF5fBzMho9UEuW7as3zI1z6MeWSyMCtKjSnrWhgSyCliie2FQFqNVq1aoVasWOnTowD0YQqEj5t8rFB1SHTnLbKBULLHIbKH/YyBuuOEGrviD0QQM0oOvD/LdKixZwaCHNdfpdGLp0qUA/C3IwiA9pbmplcDvqPlWKJvNhueee052G37uZwa/PLZesPPL8t2G8gVIzJr73Xff+V3jhx56SFaUKEEu3Z8a5LJYsO/4IlEPCzKz5Entq06dOlzfKMwHzEdYTEEpzILMxJXYb3/xxRcBAG+//bbfd8KgNin/W7vd7vPsscDUSZMmcZkqpHL2BkIPofb4448D8C1FL4T/vOjx7ARTSW/OnDkAil++5SrmSaGXCx2leSPUQgJZBSwYSWpqzePx+A0erIoQH2aFYAh9s1gKLn46I6DYQnXlyhVkZmb6DVBiVmLmJ6e2UEkwAT16I+uDzPtcUHJNqqh8GdCKHgLZbrdz7izC/bGXoZycHGzYsEFXFwv+foQuFoE6cTa9yycUHX+wPshiCKt3MeEjJVqFx+7SpUvQKex69eqFdu3aBbUPQP78XLp0CYCvhXXLli2S1TaVwizI33//vWTmDubPKuaXzHLwqu1bZs+ejZ49e2Lnzp0+LmhyLwfLli3z+05YtEOqHVKCbNeuXZxw5gflKQmyfu+99wDoI5CViEW+wNdLXGqtpNewYUPOL37atGn46quvVB1bl0IhQW2tHRLI1oYEsgoC+eYy6y5fgJw4ccLPr1g4UAuDOVhKK2EWBKfTyQ1+QgsyC1ARmwJXK5ALCwuDno7VCyURzDO//x7vL1kCAEjr1CkczdIFh8Mh6UfILMjNmjXDiBEjdHWxkLMga+nE27dvj+XLl+vSNkagPL9a4N/TXq+XE4xS93ooos5HjRqlOViJj9z5YTNdwhf5YNOt8UWSVG7fxo0bY9++faLfMV91tQKZHffAgQM+L2hi10e4jC8UlQYp8tMg8jl06BCeeuopAOrLqTPr9fr161VtJ4YSwbukpD9Uun4ggvFBZgKXBb7yi7Io3d6qQXpmEshGH9+KkEBWARMTaWlpot+L+QeLiWqhm0W3bt1EpyTFrIpsWk14s7NjirlwaCkUojXiXW+U5EHmf187TBH1eognh8OBkydPin6Xl5fnI2g8Hk9IfJCFFmRhnmMlqapsNptqwaBkn/z/9ebo0aPc85yamiq6jtFpmeRQ8gIhfIYbNmwY1DGZSGrRooVknATg79IlTAN32223aTru7bff7pfHVwj/ml27do0reDR48GDFwlzoYsGHPS///e9/uWXCPNtisOs0depURW2Qw+12IyUlhUtjKAbLWgGERiCrseh6PB7k5uZq7sPMmKVIKWZK80aoh66YCpiYkeqk3W63IkufsLKeVBJ/4QOlJDm/mOVDiwU5XC4WUtXIGA4564FIJ+0IUyekh3CT6zA///xzrFu3jvs7VBZk/guczWbzE8Rff/21LsdUi1Ag61F8hI/dbvfJBqDk+TMTSl4gcnJyuM9169ZFx44dgzoms+SpvRf57gjx8fFceW01x/V6vfB4PD6/N1ABpNdee427v9XEVCixWPJ/v5K+kv1mPTIEeb1eLjuS0vWDRUwgK+0D+VVltfSbumSxoEp6hh/fiph3BDAxUkFvzMUiUIfJfIwZUtsIgwH5nbKUYBDrNK1sQbbbpEtNA8Y99Hr5ICslJydHN8HGP2f83LE2m80vcEkqQ0qoCYWLBQAsWLAAQLF1laWyO3z4sOgzYlWBLJYaT48pZv5UuRqBzO8vL1++rPpFj1kQhQL5119/FW0jY86cOZxRIz4+XvHx5CzIjHvuuUfx/oB/sjZIzVaowe1248SJE5g9e3bAdcuVK6e5gp1esD7k2LFjml4Q2DUPViRHuosFoR7zjgAmRiptGis1HUggs0GaIWVBFi5TYqkQKxEtly9TjHBakAOhNEiPES5RE2oLsnBaeseOHSFxseBjs9nQsmVLLvuJkQgFoF6DC//ZZdkrNmzYINsGMyL3AsHuHX5RCK/XG/SzodWC/Pbbb2PFihXc31oEstfrVfQbhN8vWrRIdHmgfbjdbtliLnKlreXYvn27pu34qHE3cDgchrtYAP8Ye8TGJ6XHD+Z3mCFPvpn7E0IcQwTye++9h2bNmqF58+YYPnw4CgoKcPz4cXTq1An169fH0KFDTVWsQojUjR4oBdvBgwdFl0+dOtUvY4UYagVyr1690KBBA1ELglRbgPBakPv06SMbtBFoulN4JcqrzHerFT06OzmhMGHCBJ/gzfT0dNUvOlLICWTgnyBRIwmVBRkAatSoAY/HwwlIqfvLChZksTYy8cLvL4TWV63H9Hg8OHLkiI/7RiDKlSvnU4xCS5CeVoHMggnVuOiw3yl3rKNHjyren96oEYr8/nPEiBGajxlMkB7wT1+n9ZkK1s3CDEF6Rvcnw4YNky23TvgT9it25swZfPDBB9i2bRv27t0Lt9uNhQsX4tlnn8XEiRNx5MgRVKxYEbNmzQp30xQj5/8rZ0EWS5gOFE87BvKnA/4ZWORKV/MzX3i9Xq5QCR+v14vu3btLTr2FM83bnj17sGfPHsnvA1bSM4hQW5AdDoff9Zk2bVrQx2T71tqucBHKwYyJhkAizwznQQq5FwgmJPj+43q6WADqC17wz6VWC7ISkS+8ZvPnzwcAJCYmKj4esyCb9fqrEcg2m427VlpyEPP3E4wFmcXPaM2OpEegntGV9Izmqaeewqeffmp0MyyFIT2Ay+VCfn4+XC4Xrl27hpSUFKxfvx6DBg0CUBxQwQoomBEp66rQgvzjjz/i0KFD3PdKH5bly5eLRigz0SrVcTdo0MDvGFlZWX6ldZklUizjBfveLGneZF0sDAq8AIwRyGIp/LQg1Xa2XC9XjmAIpQVZzse0WbNm3GdhOkYzIZfTmP9csN+ph4sFX6So3ZfNZtOcB/n69evYvHlzQKsuOw4ftRUUgeLf5nK5fPbFcsoL+eWXX1TvXyk1hwzB3+npKHK5kMt79pUYUxh2ux2dSlJfBhOkGawFmeX+F7qOKUUPC7IRUCU9axN2R9Nq1arh6aefRs2aNREbG4vevXujXbt2qFChAtdxVq9eXbR+O1BcrIMV7HC5XMjKygpb21NSUlCvXj1kZWVh165dXLq36OhoFBYW4v3330e1atVw4cIFAEDVqlWRn5+PlJQUAMCFCxe4z/x2C5fVqlULEyZMEP1tKSkpSEhIEP0uMTEReXl53Hdly5aF0+nEwYMHfdbfuHEjUlJS8O2332L8+PF++3E4HLDZbGE9t1LHuuL1omJiIi4IAhsBoFxCAgrtdu78AeDW8xYWwhmi9qekpKB8+fJBn5+EhAS/a8+smmyA5v+2mJgYXa7J5cuXffbLyMvLQ0xMDKpWrcr55/KPd9NNN+HYsWOoXbt2yO+NgoICpKSkcM+PHuebUblyZeTl5WHs2LE+Vd3atWuHli1bon79+vjjjz8AFP9+Ne4E4SI3NxcpKSnIzs72s+baec/E+fPn4XQ6UalSJVy6dCmoc3j9+nWUK1cOKSkpuH79uup9sWuZm5srWwVOyKlTp5CSkoLo6Gh4PB5kZWWhcuXKAPz7jStXrvjc2+XLl0dKSopkW8X647i4ONhsNiQmJnLtjImJEX1mypUrp/g8cONAYSFsIv2Z2Pp7z5zBvF9+wdrt27Fi8mRcPnQIv/76q2i7xY5VqVIlztghdx4CcfXqVVSpUoXb/vr16yhbtqzP/uSeE9aevLw8TZZg1vaYmBi4FJ4/PklVqiDP4wnZmCBFXl4e99svX74c1jEVkL8mRGDCLpBzcnKwbNkyHD9+HBUqVMDgwYNVlR4dN24cxo0bB6C4I0tKSgpVU/3IzMxE48aNuWMyf98yZcpwHWl0dLRfm9h6SUlJPp/53yckJCj6LZmZmahYsaLounl5eT7nhFXdEx7PZrMhMzMTBQUFovu5dOkSYmNjw3pupY511enE+XPnUJkXcMS4nJuLGK/Xx++arectLERUiNqfmZmJwsLCoM/PtWvXRK9PUlIS4uLicPr0aZ/f9tFHH+lyTWJiYkSDZSpUqICkpCRkZ2dzKdD4x1u4cCHWr1+vW7lkOeLi4pCZmYmyZcsiMzMT169f1+1+zM3NRbly5fwqerEX79tuu83vuoTzWVACOy+VK1f2K5Wdn5/Ptb9SpUrcixW7vsEcMy8vD5mZmWjWrJnqfbE2ValSxS9DjxxOpxOZmZnIz89HVFQUkpKScOHCBXTv3t2vDfw+DygOOL548aJkW8Wev8LCQhQWFiI7O5v7Pjk5WfSZqVq1qmJr+ogRI/DOO++gcnQ0bCL9mVjbnuIFA1aOjkZ+yblgBPpd/II4VatWVXXe+cTGxuLs2bPc8aKjo3Ht2jW/4wdqT2JioqbMOOfOnUOlSpUQFxeHIoXnj8/58+dRweEI+3Ps5Y1PwT5/WjFb32Ulwu5isXbtWtSpUwdJSUmIiorCgAEDsHnzZuTm5nK+sqdPn5as1GQ0Ysn22dTJPffco2kac9CgQXj11VcVry8VCSwMaJOaVmJtZEn0hZgpzZuSQiFGEOo0b3a73Scor1y5crrlApZyoZAL/GLbhUMc89ugV4on4b7lrFiTJk3S7Vihgl3DQPc/38VCryC9cuXKoXz58pr3oyVID1D2G4T37sWLF9U1ruR4Qh9kqeOq6Qd69erl48KjBbHnYMGCBT4V+tjL7fjx433aHUyfHqwPsJj1XQ26BOkF1QJtUJo3axN2gVyzZk1s3boV165dg9frxbp169C0aVN0794dixcvBgDMnj3blNGWderUkY0Ejo2NFX0Ixo4dy32eOnUqbr/9dp/v1VYYksrDvG/fPpw+fZr7W6pDYW2sWrWq6PdmSvMmJ2aEv+6hO+4IfYNKEOay1kIgH2RWVhwo9gvXq4MNlMWC/d+tWzddjqeFUKV5AwLnuZUKpjUTySUVI8WuJf+3sT5A7yC9YPalxX8ZUOZHrcd9wgSy3oJGj5RrYts/9dRTeO6557i/WQaoRx55RDcf2GB9kJnVWGsb9Cg3bXQWCxLI1iPsArlTp04YNGgQ2rZtixYtWsDj8WDcuHF4++238e6776J+/fq4ePEi7r///nA3LSBer1dUyLIbXxjYweCLzYSEBL99qE3BJPcm/cknnwRcj6W3kpp6CacFeeTIkRg5cqTk9wEtyD5/hK8D0iNgTu6lqKioCOfOnfP5W68OVspSLAzSU1NcQW+EFmS9BbLH48G9996LmjVr+n1vhiDFQDCLXCCBzESFnkF6Wi157dq14/ajBvaiqCRIT/h9gwYNVB2L7UNOIE+ePFn1PgHxwFu1iBVHAeBjGGE4nc6Q3ctq7wF2XbQETQI65EE2MKCbQQLZehhiJnz11Vf9XArq1q3LBcaYFakcx4EE8hdffMF9joqK8stnu3TpUmRlZWHAgAGK2iHXUbBMIEBxp9ClSxds2rQJb7/9Np599lkA/1if5ARyuFIcTZ8+XfZ7h92OnMuXcb2oCDEBMmvYw9gB6WFlZOd46NChft8JcxHrkexfeFyhRUsonGNV+vnpiVAY6+1i4fV6UaZMGVSsWJGrtsawgkCWc7Hgp3VkgkyPPMjsxSLcOWX3798PQNlvEH7fuHFjtG3bVtXx2O8U7isjIwPNmzdHnTp1VO2PoYdAZte9ffv22LZtm+hzwZ5pp9Opa/XNYCzIbF2tM5N6lJs2WqCaNW0gIQ1dMRWICeRy5coFFMh8nE6nX15iAKqiuqXEUs+ePTnxCxSLiqZNmyImJgbvv/++3/ZSnbUe1ia9YKK3SOScsQ7TEcKUYFLoIR7ZOWbpr/iI/Ra9rglfIIsdk32v9IUtFITCcsxgAsjlcnER/vznwwoCmQkNsfPD7x/4FmS9XCy07kurwGGuYFevXlUtkM+fP6/6ubHb7di7dy9XTIlfunvv3r2aZ4/0cBNgPsxPPPEEAPGxgF8oxm636zIbGGweZL2qOGrFePsxYUXMoYIsgtgUX9u2bVUJZGZB3r17N9auXQsA6NChA5566ilV7RBD+JbNXEKE67N1JH17wyiQd+/eLVu5jYlfOX9qJqLDaR/QQ0Sxc3z33XcrWl9vH2Th/oQC2chAzVDnQWYlk9lvvHbtmt+xzYzNZsORI0cCWpD5Psh65kHWcl127Nih6bis5POPP/7o8xv4xgAh/PzEWgRyjRo1uL+FVs9jx46p2h9DjQW5QKKSLLuedrsdTqdTtA9nx4iKioLNZkPXrl01u4Uw9KqkFwxBWZANCtLjY7QFm1CP+UcCEyFmQbbZbJwfo5LADtZJPvbYY7jnnnsAFKfdUtOBqBHIUVFRfp1yIAuyHtOxSklLS+PyScvhEekc+YMFEN4OSA8Rxa652LUXs96GyweZpQ1r2bKlLsfTQqgr6bndbng8Hs6CvGTJEu57swSoBkLKn1PMxUIPFxW+Bdko2D37zTffcFZUPqxvY77HdevWVX0v2Ww22Gw2n6qkfAYOHKhqfww9BDK/UIuUVZW99LF19Ag4NIMFOej7zgCBym8zCWTrQQJZBWIC2ev1YsmSJXj66acVBVKx0s/8DkNJ8AkfKYEsFsggZ0GW6nDM5GLB/I4lp9dsNkMEsp6V9MQEsliAnN4CWcqCzNpjJguyntd2+/bt2Lp1K9xuNyeQ+b/VCi4Wcki5WBg9za0H7D5ISEgQfZHhvxy0bt1aU8lovqjs2LEjbrvtNp/vtcYfqDl/crN7bF9y2VhSU1MBFL/sFRYW6jJ7oIcPslaCdrEwQZCeWcZUQjl0xVQgJmTtdjvKly+PxMREyc64cePG3GeHw4GioiKfB1atxVaqo8jPz/fxZWYWZCHs2GawIAfC6XAgsXx50UwWwi4vnC3WMw+ymCATW6Z3Byt1jQOVNA8HoQzSA4C//vrLx8WC/5xYfSDji0TWV+jxTPOzWISzf+Bf+0DXhh8ALVYyWgn8IL2lS5di2LBhPt9rnWFQY0GWytwjFMhiYwF/nGICuTRYkINO8xbU1sFjljGVUI61R4IwI5XFAiju/MSC74Di1HYMMZcHtdYdqY5iw4YNeOyxx/zaJbW9XJCemR5mh8MBl0RbbQAe7Nu3+HMY26ynD7LYtRdbFq7fJ9eucBHKWYEWLVpg+PDhPi4WpcmCLJXmzSx5kNWiRiAnJyejadOmAIp9nnfu3Km6rb/88ouPD7MQrfeHGoEs5lIG/HM9HQ4HHA4Hzpw5I7oOayMTlnq7OKi9n/T0f9eC8fZjEshWhASyCqR8kAF5a0XPnj25z8yCzEcvC/LNN9+Mhx56CKmpqfjjjz/88jZLuVoIMZ1ADpALeTCLMreoD7LYvuTSCeqFlIuFGQhlFouqVav6ZbHg5+IurQLZ6DzI4SA+Pp4LfgaA48ePq/7dhw8fRm5uruS9p/X+YGL1bE4Oag4ZIrtuIBcLm80Gh8Phk2GDvy3/JTfY1HLseEa6WIgdX8s+wo1ehVoIYyCBrAK5N3FmHRB7CHr06IGMjAwA4hZktW/4Uh1FixYtUKlSJQDFkdZCFwsWqc86XymLt5l8kAHAKSGQuelGA7JY6NHZ/fDDDwCUu1iQQNYHJhr4Psi1atXivs/KytL9mOGE37/onQc5mDRvWuH7/Grpl/Tuy5iLxaxZs1Rtx8aIi5cvB1xXasaMS23pcEj+Lv61ttn8y2ZrQejqZLUgPaNe6hITE7Fs2TIA5upfCWWYRwVZADkLMnOxUJIHWWhBVjvgSFkE+AF57I1bTGgFSvNmJh9kQN4KYuMF6YVT1OtxrJycHMl9hUMgh3v/agilQGbPicfj8Yn4Z0i9OFoFYXwDW6ZnoFQ475Vg/cP1bit7NhMTE1Vvp1Ssbj90SHQ56weZBVkMvsGFCWS9zoFWoWm4i4WBs6IdOnQAYK7+lVAGCWQVBHKxUNIROZ1OuN1uP38uNR3IZQkLhNDHLVCQnhnyIK9evRqrV6+WXcfpcIi7WBhoQdbz/CgVw5EkkLdu3QogtALZ7XaLBiQyH1ar8tJLL3Gp+s6ePYtr167pIhCMcrG44447uM/hsCBPmjQJN910U0AXC7X7jYmJKc4ooeA67Dx8WHQ5mwWUMn4AvuMUu2Z69Fd8NwerWZAB44P0COtBAlkhXq/XJ/iBwQTpr7/+ii1btigSyEILlZ4dmFAg8/crFMZyWSzCJZBbtmwZMN+uXHALy1nKPocLPXPlKvVpDJeLRbly5XQ9jhb4FrAyZcpw1dT0gD0nbrcb2dnZ3DJGTEyMbscygmbNmnHFZwYPHoxnn33WFJX0tMJ/yddyXC15kOWslVoFMnsxU9KaQsEsI+PRRx/1aacYwiA9vSzIQj/gcAbplYZKemYyQBDKIIGsEOZ2ILzJN2zYAAD46aefAAR+CEIpkB0Oh2glPcbhEquEEhcLM+Gw2+GSSfNmRB5kPQO5lO4r1Gne2N/Mj91IevToAaC4Tb///jsWL16s2775FmR27tPT07nvzeR/rwfnz5/X3YJs1GAfDgsyP82b3P746TuV7tftdisKJt57/Ljocn4aT6l+49VXX8X+/fsB6OeDzPZlpAU56DRvJFAJlZSukSCEyKV4A4AuXboACNwRSAlkPSgsLOTS/rhcLj8L8qESvzYzuVhMmjQJkyZNkl1HLouFDca4WOghkFm1PKUdd7hcLBo1ahTS4yiBf36TkpJQsWJFXffNLMjsPm/Xrh33fWkTyIA+L+FGpXnjYwYLMgBkZGRIVjKU26/Stly7fl32ezkXi02bNnGfA4l9NRhpQQaCGyfNkHmFBLp63nvvPTRr1gzNmzfH8OHDfV4Qw0HpGwlChM1m46YtxRg2bBjq168fcD9MIAfjgyzFBx98gG+//RYA8MILL8Dr9fqIcRYcGKhQSDgtRPPnz8f8+fNl13EESFXE+fSFsQOqUaNG0PtgwRtKCZdAfvTRR3HixImQHktpW0IhVsUEMt9lpjQMZEJBoGeQnpFiQ8tvUHsPMYEcqgwqSpASgy+//DL3WclLerDBbcJ9abUg63HvBYPX6zXcB1ksHoiQ5syZM/jggw+wbds27N27F263GwsXLgxrG0ggKyQqKgrvv/++7DpKOlUxC7Iagaz0IWNiuF69etwyNo2sRCCbyYomFaTHryoFhE/Y7NmzBzNmzAh6P1osW6GEL0qNLDPN2gCELs0bE3rseTKDhUlPxASyXtPcpd3FQk9RKdoW3rX568gR3PH8837r9bvxRtHt+c+lkt/F3Dr0drFgfysl0l0sDhw4YIrYDqvhcrmQn58Pl8uFa9eucSXUw4V+kUYRjs1mQ35+vmYXC6UdiJoUVMJpuLNnz3LHY9+LYbY0b0qD9JREh+uB2vROUhgtkM10jYWE0oLMF3ppaWmoWbOmT65dxhdffKH7sY1Cz1LT7LMRhCPNm55uCWL75ve6m/fuxa6jR/3Wqy8hBPh9tlI3r1AF6andNhj0EMhGUr58eaObYEpcLhfat2/P/T1u3DiMGzcOAFCtWjU8/fTTqFmzJmJjY9G7d2/07t07rO0jgawTp0+fRmZmJqpXry67npiLhRqBrMYSJLT0MHFtNQuyZJCeIM1buASyXigZNOrWrYuRI0fi9ddfj6g0b6G0IPNfuBwOB/71r3+JrnejhBXPqughkI32QdZbIN95552oXLmy3/qhFMgenpuCQ+L3SMVc8F0clAjk0mJB1sWqb6L+jSjG6XRi27Ztot/l5ORg2bJlOH78OCpUqIDBgwdj3rx5GDVqVNjaZx4VZEHGjx/PVUNr0KABqlSpojmLhZrORs30N3+/7O3LaoVCZIP0bP8UCln955/hbFbQKDnHx44dw/nz5wGEPnjMTNc81C4WLBe53P6tnu6NoZeolZvJCRd6+yB/+umneP311/3WD6WLBevL3B4PnFLV8CQstexaVq1aVZWLhd4WZLWW5GCDmoOxXgOlz4UqEli7di3q1KmDpKQkREVFYcCAAdiyZUtY20AW5CCYPHky95lV0mMJ+qWw2Ww+A43b7cbRo0cVd2Dffvut4hy8wk4hISEBwD/CWC6LhZnEklSQHpfmraSt2w8eDGOrwsfnn38OQP8gDzPNEggJtYtFoHy+rDS8VeE/+wcPHtTleRZ7uQ83an6Hmsp1wmOEqg/kVwV1y8wcyvXN999/P2rWrKnod4WqUAj7W822wRB0HmQTBOkR6qhZsya2bt2Ka9euITY2FuvWrfNxxwgHJJB1Qo3fmtvt5qrhsfzJSjuwm266SfK7WrVqcYF4/fr1w6lTp2RdLKQGu3AK5BYtWgRcxyFVSQ/Fqd3MJOb1Jioqigu4DLVF00znMZRtMargRThh8QYAkJWVpVue9aKiIs3nrGzZsrh69WpQbVDzO6KiojRZT0MdpOfmxYBocbFgv0eNi4XRFmQ9XCyCrqRXSp/10kqnTp0waNAgtG3bFk6nE23atOH8k8MFCWSdYIOH0o7g4sWLAP6xFOgxgL300ku4//77ARSLKaEAGDt2LAYOHIjbb78dgHQnF04f5DVr1gRcxxkozZuJLaFyKOmwy5Urh5ycHAChTxNkpgGEXdNQBukB5vrNesKPhYiLi0NhYWHQ+2QWZK3n7MMPP0R+fn5QbVBzbJYz1Wxp3twsBkTOgizSN+dfv+7Tpyu1IIdKIFspiwU5WFiTV199Fa+++qphx7emsjAhWt/U2fp6dGC33XYb91nKQrZkyZKAQXpm9EFWEqRnNZRYgLxeL+rUqQNAne+5EoTX2EzXPFxp3kor8fHx3OfU1FRdziNzI9NKWlqabC55OTp16gSgdGSxYP1uoYxBRUwM5ly96mO8UJMH2eqFQvhFarRint6NsAokkHXCHsDKKQUbcPS2lMkFIVkui4WMiwVsNk4gv37ffWFsVfAoGeDcbjfn6x5JAjmUPshMNJRmFwv+bMORI0d0OY/BWpCDoX///gDCI5BD7WLBXvaL3G5J0SZ2fLfH47NcqUCmUtMUpEdowzwqyOIw64razvjcuXMA9BMnzz77LABgxYoVXKcwbNgwn3XMlAc5NTU1YPJvySA9QXR+FR3LEYcDJQMc39IZiS4WobYgm+k364nwXtHLgqzXvtTCqk7qncVCav1QBunxg6TF2paVm4vjPB9yRrTT6dOu3377TdHx9HKxuHTpEk6ePMn9Hc4gPSD4UtOl9VknQgcJZJ3Q2qkyq6BeljJW7ppvQX733Xd91mHLzVBqWgmyQXq28BcK0QulApmleQs2VVIgzHTNjc5iYXVuueUW3ffJroUR1jgtL0yNGjXy2VYp4QrSc3s8on3WPW+9hS9XrhTdXu09q7c/9enTp7l2qMEULhal9FknQgcJZJ3QavGqXbu2pu2kWL9+vc/fYvv1eDyIioqyTJo3p90Ol0yaN4aZ2qwEJVkp+OWQ9RaLZnaxCIcFuTTTsWNHLhg3JSVF13vHiFzI7OVQze9g6TC1+CCH6jcKBbJY2y5LBDKyWQ815yCUwa5W8kEmBwtCCySQdULr9CMrQalXB3bt2jXus5wPstPpLBU+yPxfZyaBp4SePXti6dKlsuuwlxkgsgSynsGrQiIhSA/4p0/Sa4odML6CnhaBbLosFgFcLKRwa5j10Ps5qlmzJgDr+SADFKRHqMc8KsjiqLF4ffbZZ9xnvaeS+UVEAglkq1TSy79+HXlXrgRcz0xtVoLD4UDHjh1l1/F4PIb6fhpFKC3IkRCkB4RGIIfazUcKLfeDVguyHjl3pVDiYiEXuGekQG7UqJHm6290qenS/jJMhAYSyDqhxsKxbNky7rNeZWAZN9xwg8/fWgSy2SzISzdtwvy1a/2/EHR6VvNBVoLH4wlYnVErZhaH4XKxMPM5CBYmZvSqpAZAcRVPvdHiYsFmXsyU5o0/G+Zyu1X1WW4NAlnP54hfAVat4Fy1alVQxyYfZMIIzKOCLI6abBT8alJ6C+RBgwZxnw8cOMAl5ueXaGR+rVbJg/z4gAEY3K2b6Hf8dppJ1OsF34KsN2a2qoSjUEhptyAvWbIEAHRNzWbUMxaMBVnt87N161auP9ebY8eOoaCkaIvafnZPerpmC7Ie103o5qCmHSzQWCt6BE6W5medCA2lT1FYAH6HrbdIEebKZQKZL8o9Ho9pLMjTpk3DtGnTZNfhT0vyEZ670tb9jRgxAsOHDw+ZBdnMAjkcPsih2r/ZKE0uFmr6pZtuukn1NgDwzTffAAjdvTF3wwYA0kF6UsfduG+fZS3IDz30UFDHDsYH2cz9HGFuSCDrRLNmzQAo64gOHjzIfdbbgiyEFSLhdxJKgvTCxahRozBq1CjZddZu3473Fy8W/Y5/3mylzIL8n//8B++88w46duyIp59+Wvf9R+rAESlBemPHjgVQugSymt8xYcIETcd67bXXVB9LDayMNL/U9NWSsthytKpTR7XVWU8LMl8g8/ethOzs7KCOrYeLBUGopXQpCgNR4+/GX4c99HpaCVmkMX+//A7S7XabxoKshL3Hj4su90vzFvqmGEJcXByefPJJ3fdrhQEnFG3csGED5s2bV+pdLNgzrKdANqpf0OKDzFB7D1WtWlX1MZRSpkwZ9G3XDkCJb3jJdTnLE5BS16pm5cqGWpD56e/UntOVEnmd1RybLMhEuDGPCrI4ajruF154gfvs8XjQvn17xMXF6daWxx9/3G9Z3bp1uc9ut1s2D3I4fZDnzZuHefPmad7eymnejMbj8SAhIcHoZoSdZs2aoWnTphEjkF0ul27C1mgLspbf8d1336lav0qVKgBC05+0atWKm+niu1gwv2Tgnz7tmeHDAQDN69QpXl+Q5o3NEIgRHx9fvC8dXZWEs45q9hlsFVC5wlaB8ILGBkIbJJB1gj2ASt5y25VYEIDQWGv5oqdBgwZ+y37//XfT5EF+5pln8Mwzz+iyr9KYxSKUmG2mQIxQlNeuU6cOGjZsGDECGdBPIBglkINJddi5c2dV64fymbDZbPCWjBGLfv6ZW853seBeBkp+6+znn0ePtm39/OaZC8kbb7zhY/Xu0qULvvjiCwDF5aH52wTD1q1bucBPtQR73zgcjqCC9ErvU06EEnOPjhZEyVsuvwPWMwUTo0uXLtznSpUqAfCdZrp8+bJpLMjB4BekZ4E2mwmzX+eMjAzExsbqvl/+YGvm329GjBLIWnMaA0C5cuXCdqxA2O12uEruvZW//84tv8YTyOy4Dt644CgJVOa/1LJrce7cOZ9xh7/O6tWruePqwc8lol7ty2VycnJQxyUXCyJY+EXUlEICWWeUCGR+xxIKgSxmdeMf86effvILuOATijaFCp80byR2VOF2uzFs2DCjmxF2mC9laR84+c8wC9YNFqMFspZ+Sa3/aygL8/D73UtXr3L34BVeeWnWj7HfWiY6Go4SYc1/qU1MTAQAFBQU+PTlYi++ev2WO++8U9M+H3zwwaCOe+XKFeTl5Wnenl6EI5ctW7agadOmaNy4MQBg165dGD9+vKJtraGCLISSt1z+IPP5559jy5YturZBbBCrV6+ez99yFmSrTD37CRwLtNlMuN3ukGTHMDuRkgc5FALZqBfnYMqtM59ipYTyJcDpdKKIJ2b/Tk8H4CuQ+ZknTi5ahPi4OMl7du7cuXjsscd8+nK+BXnMmDE++wyGu+66Cy1atOCOoYa///47qGP/+uuvmvuq0v4iTMgzceJErFmzhnuhbNWqFTZu3KhoWxLIOqPEgszcHgBgrViFuCARsyCPGzcOx44d4/4OlMXCTMLhNplyzPx2lgvBdHxpJpRFSMxMpKR54z8bly9f1mWfRgnkYCr4qZ2aD7UFmS+QZ5Skr8y8eBH7TpzwOS5/RozvYsFvV48ePRAdHe1jXeXPADJRYHShkMGDB/uMe+GEgvSIGjVq+PytdNwjgawzSgSysJiH3oh1hna73SeVnFmC9JRwoSTQRAhf3vzx2WdoVrt2WNpTWnC73REvkEvzwBlsai0xjCo1za6TlkwGlStXVrV+KJ8Jh8MBl8hveO/bb3FbSbCyXcIHWeqljrU3NzcXgO8LgdZy21Jt15rmrWnTpti7d2/QbdBK6X3KiUDUqFEDW7Zsgc1mQ1FREf7zn/+gSZMmirY1jwoqJWhNRRNu5CzIZgveqp6UJPkda2WyQdYJK1NUVGSqF6FwESlBevwZI70w+oVKrTV4wIABePnll1VtE2qBLFYVlI9Yaja7SJAe/zsAuH79OgCIBvLplQdZqwXZSEr7TBEhz2effYaPP/4YZ86cQbVq1fDXX3/h448/VrStMeaAUkw+z5fMzERFRZnCgpyRkRFwneE9euB/v/4ahtZELiNHjjS6CWHDZrMFlTIqkjH6hUqtAeKjjz5SfYxQZ7GYzUvvJrUO4GtBXrJxI/Ly8lCvc2e/djERXFRUBMC3/9bTgswXyFYTnVYR84S+uN1uTJgwAfPnz9e0feSZj0IMm+YyO3J5Jc1mQe5QEn0qxGqdtJnp06eP0U0IG/wsFma6z62AkRbkyZMno3r16iE/TihfAi5cuBCw3xJmsWCs3bULgL/YEwpkvg9yMNk/hARTajpYunfvrnlbGiciF4fDgfT0dBTyCvGogSzIBlO7dm2cKAnO0JvHHntM8rtAWSyMthTxcdjtnNXPr10kcHTBKq5BeuBwOOAVVCUjlGGkQFaamilYQmlBVtLXi/kgM8TuWX7FRMDXwKHnb+EbVcItOn8OYHUPBD3nkUvdunVx0003oX///ihbtiy3/Mknnwy4LQlknVEaxPLLL7+gb9++uPPOOznfMb1p2LCh5HdSQXqs4wtXh8Isl2vWrJFdz2m3o8jtRgxv0CDLgH5EkssBWZC1s3PnTqObEHJCGYh49uzZgOuwUtRi96aY8UK4Hn8dPS3IwnLPRjw7zEpOEEqpV68e6tWrB4/HozqbDwlknVFqiWvQoEFxRLPLFbKOpkKFCpLfSQXphVs07NmzR9F6RW43XG43YgQp7Ejg6EMkvWxEShaLdu3aYfv27UY3w3LYZQRqWI4vYkEe2bMnKsXE4JqE+1vDhg25sYdvQdY7iwXrJ8LdX7z33nuYOHEirl69irKBV/chkvo2wh8WpHvlyhUA6iprmmcevZSg5mG02+1wuVwhcWf4+eefceutt0p+LxWkZzb/Yz4fLFlidBNKLZHkYsEPNjLrva4Hy5cvx6FDh4xuhuUIpYuFEsQEcoVy5RATFSX5Uud0OjkXC/46zCVGrzzIRvUTLC2X1uOX5ueckGfv3r1o06YNmjVrhmbNmqFdu3bYt2+fom1JIOuIWv+8UArkRo0ayXYKTqdT1HHdbP7HfD5dtsxvGXV7+hBpLhaRUihEjbWEKMZIP+tNe/aIWrCdJf6/UvesUCALs1jogcvlwrVr17hjhBP2O7T0U6X9OSfkGTduHN59912kp6cjPT0d77zzjuLS5+ZUQhZF7dttqF0s5GBWkpMnT/osN7MFmQgdkSSQHTyxQfc6ISSUlfQYU8eNE12+8/Dhf6y//EIhDgdcHo9k9cuoqCifID2hD7Iev+Wrr77C9OnTub/D+eyw35GVlaVpe3rOI5erV6/6ZEHp1q0brl69qmhbEsgGwtLmGGGxZZ2s8EYxswVZCFkG9COSBDIF6WmnS5cuRjch5ITDgjyiZ0/R5UUul2iaN2fJPet2u0WDCJmxBRDPYqF3nx7uvjeppFhUT4nzJgeNEpFN3bp18frrr+PEiRM4ceIE3njjDdStW1fRttZQQqUUm83GdWrhhg0CwrRDVrMgW6mtZkJYmz7SBHIkuFiEgk2bNhndhJATSgvy6NGj0ad1a8nv+S9tdhELspRLXlRUlGyhEK2WVyXtDAdyAeeB8Hq95IoXwXz55ZfIysrCgAEDMHDgQFy4cAFffvmlom0pi4WONGzYUFVQTHZ2dtjL/QrFwXfffYfbbrvNZ51wdnzBVHAjgaOd6Ohon78jTSCHO50hYR1YfxyKZ+Ltt99G0Q8/SH7v9Xr/sSALfJBdbrekiwU/bSffyMHWDWXqOitAz3nkUrFiRXzwwQeatiULso6onZorKCjAkiVLwiqQFy9eDACS6XrCbUGePn26j1+bHPeKVHujjk8bwvMWaVksIun36knFihWNbkLYMOIeuXb9umip6f/9+itmrV0Lt9stKZDlLMitZazWWrFK30uGlMimV69ePhWOc3JyFFeONUQg5+bmYtCgQWjcuDGaNGmC3377DdnZ2ejVqxcaNGiAXr16IScnx4imBYVWoRvOjqZBgwYA/qm6tHLlSp/vzeqDfN/tt6NWcrLPMur29COSBhEK0tNOr169jG5C2DBiVuVsdjauFRQA8LUg7z1+HABkBXI4fZAB6whkwFptJfTlwoULPi46FStWxPnz5xVta4gSmjBhAtLS0nDgwAHs2rULTZo0wdSpU9GjRw8cPnwYPXr0wNSpU41oWlBo7YTC7WIBAL///rvo9+G2IO/evRu7d+8OuJ5TwupH3Z42hPdcJLlYsICmSBHIVapU0W1fkXC+GKG2IC994w2/ZU6HA2ezswH4PqO3d+4MAPjtt9+Qn5/vv51EmrfSEqQXDNZpKREK7Ha7T7au9PR0xf1Y2AVyXl4eNm7ciPvvvx9AsS9khQoVsGzZMtx7770AgHvvvRdLly4Nd9OCxgoWZGZ94E858Am3BTktLQ1paWkB1/vu11/xsQXvCbMSyT7I0dHRuH79esQI5ObNmxvdBEsSaoHctmFDv2WFLhfnWnGpJOcwAAzu1g23tmiBc+fO4b///a/fdnwXC7E0b3r06R988AFuv/127m+rPDsUpBfZvPnmm+jSpQtGjx6NUaNGoWvXrpgyZYqibcPuuX/8+HEkJSVh7Nix2LVrF9q1a4f3338f586dQ0pKCgAgOTkZ586dC3fTgkZrhxHOjiYmJgYAkJ+fL9ppmjWLxYW8PL9lVrJimI1BgwYhMTGR+zuSBHKZMmVw/fp1ANYZ5INBzxfeSDhfDCP6l9b16+NMScaJcyWWZKDYH9ld8oyKPav8yqhiLhZ6pK67fPky55Jntb43ku5bwpe0tDTs2LEDW7duhc1mw4wZM1C5cmVF24ZdILtcLuzYsQMffvghOnXqhAkTJvi5U9hsNskbeubMmZg5cya3L73T1wRDu3btkJeXp7hN7IXAbreH9XekpKQgLi4OcXFxAHxTAOXl5aFq1aphP69Sx3MVFsKWn8+dqwu86cWKiYm46vH4LGN4CwvhNNG9oZRw+d7ffffduPvuu5GVlYWUlBS43W5TPUuh5OrVq7h69SqSk5Nx6dKlgL/bivEQfMqWLYuUlBRdrm9MTIxu+wqGUF+TlJQUlCtXLiS/k/Vp7DiMnu3aIS4+Hq2bNsX5a9dw4OxZrm8rstsRm5CAlJQUdOzY0a9dZcuWRUFBAbKyspCYmMjd19euXUNKSgouXrwYtEj8448/fK59dHS0TzvCcU0A4ALv/Cnh8vXrqFK1Ki5YdEwIBqv3XcGQnp6OChUqICEhAZUrV0bZsmWxdOlSHDx4EI8++qjfLKoYYRfI1atXR/Xq1dGpUycAxZasqVOnomrVqsjMzERKSgoyMzMl/ebGjRuHcSVViOLi4rgE4mbglVdeweTJkzkrbSDatGmDlStXwmazhfV3ZGZmAgB3rvnHzsvLw8GDB8N+XqWOVxQdDVtsLNfmyrGx3HfZFy6gnMPhs4zhLSxElInuDTWE+9xnZmbC6/Wa6lkKJTk5OcjMzITT6URCQoKi323lcxMXF+f3nGtl9uzZAMxxPkLZhszMTNSuXTskx2B9GjsOw1ZYCO/160BhITIzM3EyMZHr28o5HMi9cAEpKSkYOnSoX7uY9bhChQrYs2cPKlasiKSkJFy+fFl2PFWD1+vl7iOv14uioiK/doT6mgBAZd75U4LT7cb58+dROTrasmNCMJjhWTWCIUOG4H//+x8SEhLw119/YfDgwXj++eexa9cujB8/XtRVSUjYfZCTk5NRo0YNHDx4EACwbt06NG3aFP379+c639mzZ+POO+8Md9OCxuFwKBbHANC4cWMA4Q3SC8TmzZuNboIovdq3F11OU2fBM3nyZEs+b1phEc1mdSfSm9dffx07duwwuhmWI5xuR1EOB+x2O1y8HPUJJTN8QElqQo8Hf/31l+h44XA4UFRUxKXMFCs2EixlypTx+Tvcz87AgQMB+PpmE4QU+fn5SE1NBQDMmzcP9913H5566il89dVX+OOPPxTtw5Ds4R9++CFGjhyJwsJC1K1bF1999RU8Hg+GDBmCWbNmoVatWli0aJERTQsrzD/MqEF6xIgRftkszOqL+nd6ut8ya3nBmZfx48cb3YSwwp67q1evRoRA5rtTEcoJp5/t0QUL8Nrs2cXpB0uWNeeVw3WWVNLzeDzYvn072rZt67M9My61adMGwD9jip73N39K2ggf5Dp16gAAFm3ejAcHDVK8HQXpRSb8e3T9+vVcYJ6al8aAAvncuXN44YUXkJGRgVWrVmH//v347bffuCwUWmjdujW2bdvmt3zdunWa92lFWLlDoyzIKSkpSBbkFjZrAYUmtWrhtIj/GHV8hFqYQM7NzY0IgUxoI1zGgnIl7gIOux0ut5sb2KvyirLw01zK9dE7d+4E8I8wrlmzJr777jtd2tm9e3fMmzcP2SXBg+F+dho1agQA+M/SpaoEMkAzjZHIrbfeiiFDhiAlJQU5OTm49dZbARS76ijxPwYUuFiMGTMGffr0QUZGBoDicsozZszQ3mqC48KFCwCMe3idTieOHDnis6x69ephbcPq1auxevXqgOtVTkgIQ2uISCDSy+4SygiHlbRtgwb45uWXARS7Sbg9Hsz76Se0adAAg7t149a7nJ+P3SWzaEoMKnwLcueSHMrBcuONNwIofrE0woLM0oFeL0lnpxSrZdwg9GHGjBkYMGAAateujU2bNnFVJc+ePYs333xT0T4CjhQXLlzAkCFDOPO00+nUJWUM8Q/hFsjz589HcnIynnjiCezZs8fnuwoVKqBjx45ha0vLli0VrRcn4ttNHR+hBeq/tFOhQgXJHOqlDVZ4I5Qs5Q3UDp6V+FhGhs+4cOjUKe6z0L1CjFBYv9mLpdIqZHrDnluPhn6fLMiRh81mw7Bhw/yWMzckJQQUyGXLlvVJEbN161YkkDVPV8LtYtG9e3cA/1iw+bhcLlNa2GJLBHKhy4Vofvuo4yNUwiwJAA2catmwYQOWL19udDPCwq5du8J6PH6uY2EgGhsjypUrh4YiBUaEVKpUSf/2lQjUlStX+rnmmRkyoxBaCajM3n33XfTv3x9Hjx7FTTfdhHvuuQcffvhhONoWMRg9SF+6dIn7HG6BPGnSJEyaNCngelElbeL735EFmdAC/4XU6GfPalSpUgUPPPCA0c0olfAFsrBvq1cSjS/VP0+bNs3HPY7/EqgXLIsFS49llWeHgvQIrQQUyG3btsUvv/yCLVu24PPPP8e+ffsUT4sT8rDclEYF6TGndX5nHG6BPH/+fMyfPz/gemyAEE6vWaWTJswJ3T+EWWA+yGIklaQmdLvdov3zmTNncPr06VA2zwerGSfoOY9cvv/+e80uRwGV2Zw5c/D1119j+/bt2LFjBxYsWIA5c+ZoOhjhy8MPPwzAuId37NixAIqd1hn33HMP1q9fb0h75Oh/000AzJuGjiAIIhj4PshCrhcWAgCKiopEBfInn3zCfeZX5wsVXq/XMqLTamKe0JdvvvkGDRo0wDPPPIMDBw6o2jagqfDPP//kPhcUFGDdunVo27Yt7rnnHvUtJXxo3bq1ocdnU2axKqoSGUn5uDhNARoEIYVVBnmi9MPSvAHA2w89JLme2IxjUUlmhxYtWnDFQkKNlZ4dK7WV8CU3NxcPPPAA9u7dC5vNhi+//BI33HCD4u3nzZuHS5cuYcGCBRgzZgxsNhvGjh2L4cOHIz4+XnbbgAJZ6G+cm5srGhlIaIfvAxxOmEBmnUdhiZXCrFy6dg37jh/Hjc2bc8uo2yOCgQZOwiw4HA5OILMZM0ZthVZhl8sVFpc9K1llrdNSQowJEyYgLS0NixcvRmFhIa5pqKRYvnx5DBo0CPn5+ZgxYwb+97//Yfr06Xj88cfx2GOPSW6n+kkqW7Ysjh8/rrqBhD+tWrUCAK7sdriJiopC7dq1OeuDUel71LDn2DHus5U6acKcmKnMOxHZOOx27DtxAgBQVlDWOaFsWdltExMTAQDHjx8PWwyJVV4uKUjPuuTl5WHjxo1cYbro6GhUKPHHV8ry5ctx9913o1u3bigqKsIff/yBVatWYdeuXXjnnXdktw34JN1xxx3cg+DxeLB//34MGTJEVQMJcZhrQzjybUpx4sQJzJs3Dy+99BKX69rMCCWxVTppwpzQ/UOYBYfdjm0ajSVvvfUWHnroIRQUFFCebzHoOTclLpcL7du35/4eN24cxo0bx/19/PhxJCUlYezYsdi1axfatWuH999/H2UDvDDyWbJkCSZOnIiuXbv6LI+Li8OsWbNktw0okJ9++ul/VnY6UatWrbBXWyvtGF3e+bPPPsNLL72E//3vf2E/dosWLVStn84LKCQLMhEsJJAJKcKdD96hYDaDlVsWwvfJDJdANvLZ8Xg8imd/aJwwL06nE9u2bZP83uVyYceOHfjwww/RqVMnTJgwAVOnTsXrr7+u+BizZ8+W/K5Hjx7y7Qu081tuuUVxQwhtfP/99/j888+NboYhrFmzRtX689euxRTeGyYJHCIY6P4hpAj3zN7PO3cGXEfqfq1cuTL3ORwC2SjRWbNmTZw8eRJuFQIZoOfcqlSvXh3Vq1dHp06dAACDBg3C1KlTFW0bHx/vc91Z5hX2v5LYL8k7LD4+HuXLl/f7x5YTBEFYHfJBJsQwoprolr17ddlPabYgjxkzBgBwgjeTSJRekpOTUaNGDS5Oa926dWjatKmibS9fvoxLly5x/9jf7H8lSPYCly9fVrQDovSxefNmo5ugCJo4I4KFBDIhhtPpDLsFWYmwVSJKS7MFuXfv3njttddw5sIFNFDo6klBetbmww8/xMiRI1FYWIi6deviq6++0rSf8+fPo6CggPu7Zs2aAbdR/JqsZeeENaldu3bYjpVaUiEvIyND0/bU8RHBQFOvhBhGWJArJyQg98qVoPcTyrZ37twZW7duBWDMs1O3bl0AwPodO9BNRR0Bes6tS+vWrWX9lAOxfPlyPPXUU8jIyECVKlWQnp6OJk2aYN++fQG3DWg+Wb58ORo0aIA6derglltuQe3atXHbbbdpbizhT8WKFY1ugg9m7kya8cQ7BV8QwUIWZEIMLblWg+XpoUN12U8o7+kaNWoAML6i6f+tXq14XRonIpsXX3wRW7duRcOGDXH8+HGsW7cOnTt3VrRtwCcpmJ0Tyujfv79hx05KSgJgfIenhImDB6Nnu3Y+y8ws5gnz0qFDBwBQlS6IiByM6A+jFFh+jXaxYMFS6enplup7rdRWQl+ioqKQmJgIj8cDj8eD7t27K7ZIBxTIweycCMwTTzxhaNlulov5448/NqwNSln9xx94f8kS7m+yDBBaUVOqlCDCgZI0b0oIpYtFvXr1ABhX3EoLNEpENhUqVMCVK1fQtWtXjBw5EhMmTFBsGAn4JLGd33zzzRg5ciSqVKlCVhcdeeaZZww9PutMrVAk5O/0dL9lZBkgtDB69GjTuTYRkU2RgqDA/fv3B1wnlC4WcXFxAIDr169bqu+1TksJvVm2bBnKlCmD9957D/Pnz0deXh5eeuklRdsGFMjdu3dHXl4e3n//fcybN0/Vzgnzc4xXuhkAGjZsaFBLCCJ8VKtWDQ899JDRzSBMSuPGjXHgwIGwHvPvkyd12U8oBTIr7JSVlRWyY+gNzTRGNsyge+nSJdxxxx2qtg0okF0uF3r37o1KlSph6NChGDp0KFf3nbA+iYmJuHjxIvf38OHDDWyNOqjbIwgiFFjJOiqkTJkyIT8Gf8wwO6wwBBGZfP7553j55ZdRpkwZ2O127n4QGgfFCPiq+fLLL2Pfvn34+OOPkZmZiVtuuQU9e/bUpeGE8bz55ps+f/fp0yesx582bRqmTZumaN3RvXuHuDUEQRDA4MGDw56t6YZmzQAAdouIOSuJTiu1ldCX//znP9i7dy9OnDiBY8eO4fjx44rEMaAiD3KVKlWQnJyMxMREnD9/XnNjCXNx/Phxn7/DmQMZAEaNGqV43bk//gjAt2QkQRCE3jz88MN4+OGHw3rMTk2aYPeXXyK+xM+XkKfQ5UK0goBEGiUim3r16nG+82oJeHd98sknWLRoEbKysjB48GB88cUXikv9EeanVq1aRjdBNUVuN9cxkmWAIIjSQoVy5YxugmKM7nuvFxYqEsgABelFMlOmTMGNN96ITp06ISYmhlv+wQcfBNw24N116tQpzJgxA61VVK0hrEOzkmk9o5g3bx4AZZbk5nXqYO/x43h+5ky8M348QBZkgiCIiKJJ9er4+/Rp/LZ/P3q3bx9wfZppjGweeugh3HrrrWjRooXqANaAAtkK6b8I7Ridso+luVMikN9/7DH0ePJJfLthQ7FAhvFWDIIgiEjEqL73iTvuwEOfforzOTmK1vd6vQCNExFLUVER3n33XU3bUp3VCCclJcXoJihGWGmK7AIEQUQSlLv7n3Fgwbp1irchQ0rkctttt2HmzJnIzMxEdnY2908JoSu5Q1gCfsfx999/G9gSbVC3RxBEpJAjYzWdOHEi3nvvvbC1xSjR2b6kmt8ehZkIiMhmwYIFAHy9IZSmeSOBTHAkJCQY3QSCIAhCggEDBkh+d+XKlTC2xDgSNLgFkiElchFm6lIDCWTCMqRWrsx9Tj93joIvCIKIGNa+9hpqDB0q+f0XX3wRxtYY77bwQN++itajcSKyKSoqwqeffoqNGzcCALp164aHHnoIUVFRAbclH2TCMvBT+pwt8SEyupMmCIIIBw1SUhAfHy/5fbgyTY0vCZA2mv/+8IOi9aiSXmTzyCOPYPv27Rg/fjzGjx+P7du345FHHlG0LVmQCWRkZBjdBNU8MG0ayoahpCpBEIQVuPPOO/HXX3+F/Djjxo3DJ598gry8vJAfSy9IIEcuf/75J3bt2sX9feutt6JVq1aKtiULMmEoGRkZmgR63tWrAKjjIwiCAMLXF0ZHRwOwjuuCNVpJhAqHw4GjR49yfx87dgwOh0PRtmRBJizFrGeewf3TpgGgjo8gCIKhtgiCVpjv5pYtW8JyPD0gM0rkMn36dHTv3h1169aF1+tFeno6vvrqK0XbkkAmLEXHJk18/qaOjyAIInwWZGdJLEh+fn5YjidGoxo1cPDUKUXrWsXSTYSGHj164PDhwzh48CAAoFGjRj4lp+UggUwYSp8+fQAAa9asUbQ+P8WPx+MJSZsIgiCsRrgEMrMgu1yusBxPjKMq3PIoSI/Yvn07Tpw4AZfLxfnp33PPPQG3I4FMGMqePXs0b+v2eKjjIwiCADBs2DCkpqaG/DjMlcPtdof8WFK4So7t8XgUuZbQOBG5jB49GkePHkXr1q0532ObzUYCmSjduMmCTBAEAQCIi4tDWlpa2I535syZsB1LSI0qVXDq/HkUulwoUxI0KAU5WEQ227Ztw/79+zW9JFEWC8KyuN1usgwQBEFEGA/37w+AjCREYJo3b46zZ89q2pYEMmE5tnz8MQDqHAmCICKRgV27AgC+Wrky4LoUpBfZXLhwAU2bNkWfPn3Qv39/7p8SyMWCsBzVk5JQPi4OhQYGiRAEQRDGwPyOpy1ciEcHDJBfmYL0IppXXnlF87YkkAlLYrfbDQ0SIQiCIIzBoTLnMwnkyOWWW27x+XvTpk1YsGCB33IxSCAThjJy5EhN2znsdnKxIAiCiECiSnIxpyQmBlyXHCyInTt34uuvv8a3336LOnXqYODAgYq2I4FMGMr06dM1bWe32+GhqTOCIAhDSElJMboJyLx4UdF6NEpEHocOHcKCBQuwYMECVK5cGUOHDoXX68XPP/+seB8kkAlLonaKjSAIgtCHAQMGoF69ekY3QxEUpBeZNG7cGDfffDNWrFiB+vXrAwDee+89VfsggUwYyu7duwEALVu2VLUdE8hkQSYIgggvH330kdFNUAxV0otMvvvuOyxcuBDdu3dHWloahg0bpvplicxwhKGkpaVpSm6vpHoSQRAEQZBAjjzuuusuLFy4EAcOHED37t0xY8YMnD9/Ho888gh+/PFHRfsglUFYkoLCQgDkW0YQBEFIQw4WkU3ZsmUxYsQIfP/99zh9+jTatGmDt99+W9G2JJAJS8JSvBVRLmSCIAhCBjKkEABQsWJFjBs3DuvWrVO0PglkwpI4S9L8XC8qMrglBEEQhFmhID1CKySQCUsSExUFAHBRsRCCIIiIo1f79orWoyA9QiskkAlLYi/p8OqmphrcEoIgCCLcjOzZEwCQceFCwHVJIBNaIIFMWBKW5o3yIRMEQUQeTocDALBl3z7Z9cjBgtCKYerC7XajTZs26NevHwDg+PHj6NSpE+rXr4+hQ4eisCRLAVG6Wb16NVavXq16O0rzRhAEEbl0aNwYwD8B2wShN4apjPfffx9NmjTh/n722WcxceJEHDlyBBUrVsSsWbOMahoRRlq2bKm6SAhAApkgCCKSiY2JAQBM+uwz2fUoSI/QiiEq4/Tp0/jhhx/wwAMPACi+gdevX49BgwYBAO69914sXbrUiKYRFsFOPmUEQRCEAsgHmdCCIQL5iSeewLRp0zgr4MWLF1GhQgUudVf16tVx5swZI5pGhJlJkyZh0qRJqrejDo8gCIIICFmQCY04w33AFStWoEqVKmjXrh02bNigevuZM2di5syZAACXy4WsrCydWxg55OTkGN0EzJ8/HwDwzDPPiH7vKiyELT/fb3mlxESkFBbigsh3AOAtLITTgveGGa4J4QtdE/Nh5Wsi1acFwux9WriuCf/8paSkAIDkOAAAVz0eVExMxAWTn79QYOXnxAyEXSBv3rwZy5cvx8qVK1FQUIBLly5hwoQJyM3NhcvlgtPpxOnTp1GtWjXR7ceNG4dx48YBAOLi4pCUlBTO5pc6zHL+pNpRFB0NW2ys3/Lc7GxkZmaissh3QPFgEmWS36YWs1wT4h/ompgPq14TqT4tEFbo08JxTfjnLzMzEwAkxwEAOG23Iy87G5Wjo01//kKBVZ8TMxB2F4spU6bg9OnTOHHiBBYuXIhbb70V8+fPR/fu3bF48WIAwOzZs3HnnXeGu2kEQRAEAefu3YhdsMDoZpiW2PnzETd7ttHNUAQF6RFaMU0qgLfffhvvvvsu6tevj4sXL+L+++83ukmEiSEfZIIgQkVSWhoqPPWU0c0wLRUmTULC888b3QzF0HhBaCHsLhZ8unXrhm7dugEA6tatiz/++MPI5hAEQRAEUYogCzKhFdNYkAlCDWQRIAiCIALhBY0XVkZYVC6ckEAmDKVFixZo0aKF6u2owyMIgohs2jZoAAAocrlk16PRwroIi8qFExLIhKGsWbMGa9asUb0ddXgEQRCRTd7VqwCAzXv3Sq5Tb9MmbD98GAggognzISwqF25IIBMEQRAEYTkqxMcHXKfxzz8DAGyFhaFuDqESl8uF9u3bc/9YjQuGsKhcuCGBTFgScrEIDvuZM4hZt87oZhAEQWjmxdGjAQB/HjhgcEsILTidTmzbto37x2pcAL5F5YyCBDJhKKmpqUhNTVW9HQnk4KicloZKJYMLQRCEFWnbsCEAYLYGNz3C3LCicrVr18awYcOwfv16jBo1KqxtIIFMWBKSx8HhuHjR6CYQFsSWl4fyzz0HXLtmdFMIgqOwqMjoJhA6I1ZUbt68eWFtAwlkwpKQBZkgwk+5GTNQds4cxM2fb3RTCILDIfBRrTlkCF6cNcug1hClBRLIBEFEHLaLF5GSmoq4OXOMboq18HgAADYqvkCYiKsFBX7LyO2i9NCtWzesWLEi7MclgUxYE7Ig60aZFSvg3L3b6GaEFeepUwCA2AULDG4JQRB6IqycR69yhFYMLTVNEFoheawfFUsihzMzMgxuCUEQRHCknztndBOIUgJZkAlLUsHjgRdA7JYtRjeFIAiCkMF54AAq9+kD2+XLIT/WlDAHchGlF7IgE4Yybdo0TdvVLkn6Hr9qFfJvvFHPJhEEEQjyQSZUED91KqL27EH05s24npam675jY2KQf/069/eqP/7gPov5JhOEUkggE4aiNa8huVgQhAGQ7z9hMlITE3FUwj2syT334HpMTJhbRJQWyMWCIIjIhSyhRCnDuXt3ROWplhLHftCzTqiEBDJhKPPmzdOU/JvyIBNBYfH7x37hAlJSUxG9aZPRTSFUEr92LWoMGQK4XLrv25aTg6S0NFSYMEH3fVsWiz/rhHGQQCYM5ZlnnsEzzzyjfkPq9IgIJmrHDgBA2ZkzDW4JoZaE5csBALaSOAo9sZVYjqNL7o9I4K4uXbjPL3zxhYEtIUobJJAJS8LJY5o2I4jwQ89dSCm7di2cZ84Y3Qz9CcF98/jAgQCA01lZmPfTT/6H1P2IRKRAQXqEJfGSBZkgwg89d2Gh0syZ8EZF4XRpKekdwvsmriQI7+u1a0W/zy8oAIXpEVogCzJhSWiYJvSASiYTZsVWVGR0EyxBcqVKAIDf9u/nllWtWJH7TE84oRUSyARBRBw0A0EYjoVezmIXL0ZKaipsOTlGN8UPu71Yxmw/eJBbtkokvz498YRaSCATloQ6uxDg8RjdAkIpAQR+3JdfUgBfJBKiF7+4WbMAAM709JDsX28qJyTg9s6djW4GYXFIIBOWhEvzZiErjBlxHj7MfY6bO9fAlhB6kjB5Msq/8orRzSAIwxjdu7fP3zRSEGohgUwYSkZGBjKUJnrn8cnRoyFoTeSRdMst3Gd7draBLfEndtGikE3pku9xkND5I9RggEtT56ZNw35MITE//oiqjRsD+flGN4XQAAlkgiBMh+PIEVR44glUHD8+pMchX2SCKF38MHUqAMBhN17elH/jDdgvXYLz9Gmjm0JowPg7iCAIQoDt+nUAgD0rK7THIUuoNujFwtyE6r62wPPSrHZto5tAlBJIIBOG0qdPH/Tp08foZhARRqmxHFtAsBAShOLaheq+tsjz0qZBAy6rBQC0rFeP+2yNX0CYCRLIhKHs2bMHe/bsMboZBEBiy0pYRLAQIkTwtQv1jM1OXtAxAEx58EEKziM0QwKZIAjzQqLdnNB1IUxEr/btRZe3qFsXCWXLhrk1ItDzYklIIBMEYT4i2MqmCBpwCTFMfl+EyrVpVK9eIdlvsJQaV64IhQQyQRCRi8kFRUBoACbCiUmfF4fDYXQTiFIICWSCIIox6eBHyEDXzLJY6tXG5C9i8bGxAIpdKiShZ4VQCQlkgiAIq2FywUIYRITeF63r1wcA7Dl2zOCWEKUJp9ENICKbkSNHGt0EwsyQ1cec0HUJmkg8g6HKYmGTezEww0sDPS+WhAQyYSjTp083ugmEGTHDoEb4Q9eF0EKk3jeR+rtLCeRiQRAEQRAEQRA8SCAThrJ7927s3r3b6GYQgLmmAcPVFjP9Zi1Yvf0RTEhti3RfEETQkIsFYShpaWkAgIyMDINbQkQUVp/6tHr7I5iQStcIvy+kCoYAMPalgV5YLAlZkAlrQx2PfkTS4Er3TXDQ+SO0EML75sZmzdC3c2f/LyKpXyN0hSzIBEGYj3ANajR4EkTICUdFuYUvvyxxcBO8zFE/Y0nIgkwQBEEogwb6yMYMYlMrdO8SKiGBTBBEMVYe/LRi9d9s9fZHMqG8dnrvm8QlEYGQQCYIgrAaRgkWEuTBE8prR0LWnNBzY0lIIBPWhjqe0k2orq/VhYTR973Vz19pxej7IhCRlkmCnhNLQ0F6hKGsXr3a6CYQZoQGFmWQJZkQg56ff6BzQWiEBDJhKC1btjS6CQRBKIXERmRDL0ZEBEEuFoS1oQFbP2jwsx50zawLBekRhKkhgUwYyqRJkzBp0iSjm0GYlVALQKsKTBIs1iWSg/TM3r5QYdV+JsIhgUwYyvz58zF//nyjm0GYjVAPpJE6UBPGY2WxRM+NOuh8WRoSyIS1sfJgQ0hD19Xc0PUJnkgUT3TfEBaCBDJBEOYlEkWEmaHrEdloFbh03xAWhAQyQRDFkHXHetA1sy5WCtIjiAiEBDJBEOaFBnqCUA5Zas0J9WOWJOwC+dSpU+jevTuaNm2KZs2a4f333wcAZGdno1evXmjQoAF69eqFnJyccDeNICIbMw2uZmqLGaHzY13o2hGEJQi7QHY6nXjnnXewf/9+bN26FR9//DH279+PqVOnokePHjh8+DB69OiBqVOnhrtphAG0aNECLVq00Lw9DTUEYQBkEYtMgr3ukXrf0EuRJQm7QE5JSUHbtm0BAPHx8WjSpAnOnDmDZcuW4d577wUA3HvvvVi6dGm4m0YYwJo1a7BmzRqjm0EAkTt4EQRR+qH+zXJIeRyEC0NLTZ84cQI7d+5Ep06dcO7cOaSkpAAAkpOTce7cOSObRhAEYXps4R70yRJmDcwmBo28b8xwz5rtelgE5nHQtm1bXL58Ge3atUOvXr3QtGnT8Bw/LEcR4cqVKxg4cCBmzJiB8uXL+3xns9lgk7ipZ86ciZkzZwIAXC4XsrKyQt7W0ooV/LxdhYWw5ef7La9R8n+Rx4MLIt97CwvhtOC9Ea5rkiKy7Nq1a6Z5nmKys5EEwB2iZ7xMTg6SALjc7oD7N+NzUi4vD5UAFBYWirafXV+9z53j2jWUA3D16lVD75VQX5NQnT8ASCoRS9n5+XBFRYmuw/o3ft+mpE9zXryIqgA8Ho+ubS/vciEaQG5uLq5lZYmeH7lrEnv9OmIBXLp0CXlBtktqTJAi2euFA0B2UREQ5nu2gsuFKBSfmwIDnhcz9l1qSElJ4QynfI+DUi2Qi4qKMHDgQIwcORIDBgwAAFStWhWZmZlISUlBZmYmqlSpIrrtuHHjMG7cOABAXFwckpKSwtbu0ojR5y81NRUAkJGRIfp9UXQ0bLGxkttH2e2oLPK9t7AQURa9N4y6JmZ6npwlHbvD4QhJm5wlM1ROhfs3y3lhxCQkAACiYmJk26Z3u+Pi4gAAZcuWBQw+J+G4JqE4BjP+VIqNhUembwPg07cp6dPsJeLbrvNzE+UslgoVKlRAWd5+hceQOmZMTAwAoHz58ogOsl2BxgQhzI+0UlQUHGG+Z50lL0AVK1aEy6DnxWx9l1b4HgfhIuw+yF6vF/fffz+aNGmCJ598klvev39/zJ49GwAwe/Zs3HnnneFuGmFFaOqK0IIZpl0JwmpYub+lZ950uFwutG/fnvvHvAOEyHkchJKwW5A3b96MuXPnokWLFmjdujUA4K233sJzzz2HIUOGYNasWahVqxYWLVoU7qYRBEEQBBEqrCywCd1xOp3Ytm2b7DpiHgfhIuwCuUuXLvBKPCTr1q0Lc2sIgjAzYQ9CI4jSgN7PDbO+arXCmsF6a2RfQv2YJqQ8DsIFVdIjCIKwKPQCYWEi6doZ+VsjPYOGhWEeB+vXr0fr1q3RunVrrFy5MmzHNzTNG0EEDXVApRpvqK+vRUVKyM8LETpCee3oviBKEXIeB+GALMgEQRAEES4s+lIGIPi2k4AnLARZkAlDmTZtWnA7sPJgYzYi6VzSQE0Q4SeS+hjC8pBAJgxl1KhRRjeBMDHkY0uUWqz0khZsW630W0MB9WOWhFwsCIIwH5E+oCqFSk1bl1BeO733TQKPiEBIIBOGMm/ePMybN8/oZhCRBg34hFFQkF54oXNCaIRcLAhDeeaZZwCQq4UpoIGEIAg56MWSiCDIgkxYG+qw9SOSziW9DARHJN0rhH7PS6TeN9TfWBISyARBmJdIHVAJQgv0vJgTui6WhAQyQRCmg4YTk0KWMP0g0RRe6HwTKiGBTBAEQRClAXqBIQjdIIFMEETkQlYlgiAIQgQSyARBADBXUQ6ygxGEiSDLdHCYqG8llENp3ghDycjIMLoJRCRCA35QmOllirAQdN8QFoIsyARBmA4aRhVClfQIMcwmRCP1vonU311KIIFMEARhNYwaeM0mvKwIVdIjCEtAApkwlD59+qBPnz5GN4MASPwQivGSECPUQH0LYUHIB5kwlD179hjdBMLMhHpgpYGbCDcWvOfohShILHjNCbIgEwQRidCAHxQUpEcQCqB+xtKQQCYIwnzQwKIMEqqECHq/wAS9P3qeCQtCApkgiMiDhKU2SOgQhHKon7E0JJAJgiCsBglV6xLCaxdqOWZp1xort50wBBLIBEEQVoMGe0KEUElvSwfpWbnthKFQFgvCUEaOHBncDkgolE5oUDM39NwRWoi0+4b6MUtDApkwlOnTpxvdBCISoYGLMJpIE4sEYTHIxYIgCPNB4sGUWHqqnQgazdef7hvCgpBAJgxl9+7d2L17t9HNIAASpVaCBId1oSA9grAE5GJBGEpaWhoAICMjw+CWEARBEKKUghcy6/8CItyQBZkgCPMRrgGZLGIEUboxUtyXgheLSIYEMmFtSOAQRPih5y54IvEcRuJvJiwLCWSCICIPsuxog85bZKNV4NJ9Q1gQEsgEQRRjRuuOGdtkBkhwWBa6oyMPCm60JiSQCYIgCGXQQB+ZWPmFjO5ZQiMkkAmCMB8UpEcQpYcIfc4ob7i1oTRvhKGsXr3a6CYQBEGEjYCSKULFZMggkUpohAQyYSgtW7Y0ugkEYV1ITBFi6H1fsP3R/aYK8j22NuRiQRCEeaEBxlyQNS5oQnlHm/bq0H1DWBASyIShTJo0CZMmTTK6GUSkQgM3QQSGnhNNkA+ytSGBTBjK/PnzMX/+fKObQRDWhAZgQoxQ3RdWvt9oNopQCQlkgiDMS6gHZBo0tUHnjSCIUg4JZIIgiiHRYz3omlmWkL76heq+sOD9Zr0WE2aBBDJBEOYlVAOylaeKAeu3P5KJxGsXib+ZjwVfLAgSyARBmJFIH1AJwkzQ86gNOm+WhgQyQRAEQRAEQfCgQiGEobRo0cLoJhARDCXyJ4gIgZ51QiUkkAlDWbNmTXA7oE5PP8x4Ls3YJjNh1Pmh6xI8oTyHJg3Si9gX0kj93RaHXCwIgog8yDdQG3TeCC0Yed/QPUtohAQyQRDmgwY1eej8EOGE7jdt0HmzNCSQCUNJTU1Famqq0c0gCIIgCMJErF69Go0aNUL9+vUxderUsB+fBDJBEARBEARhGtxuN/71r39h1apV2L9/PxYsWID9+/eHtQ0UpEcYR0EB/l3ysdyMGaKruA8ehK1MGcldRJ88ifJLlvgt9xYUwHH4sA6NDC+2q1dRtmxZQ45d5vvv4Y2PN+TYQuwXLwIAnOnpkvdGUPs/d654/0eOBNy/kddECueBAwCAmC1bZNuv97mLXb68+P9lywCHQ9d9qyFc1yQU954zN7d43z/8AK/Yb+AFdPH7NiV9mu3aNQCAPS9P17bHbN4MAIibOxfRf/7JLecfQ+6axJQEY8d+8w0cp04F1ZZAY4IQZ3Y2ACB+3TrYSvqVcBH9118AgLjZs+H+5ZewHhtQ/5xcGz4cnqpVQ9gi5fzxxx+oX78+6tatCwAYNmwYli1bhqZNm4atDTav17rhlXFxcTh69KjRzbAsWVlZSEpKMuz4trw8JDdpYtjxCYIgCIIoJmv1arhatgzb8WrXru2T6nXcuHEYN24cAGDx4sVYvXo1/vvf/wIA5s6di99//x0fffRR2NpHFmTCMLzly3M34KmTJ0XXKVq5EraEBPX7zstD1O23B9E6YzD6pYXw5//bu/+YqOs/DuDPE9AkKHMNpQ/Eidw0z+MODISGNmLEia4/ZEuYIo3MLVpITcjFFtofuShbTlebzpqLrVbMAoVwa9RyxeSH5ZJWaw4nh0QIKBUa9+P1/aPdfY9fJ7+8+3yu5+Mv+dznx8vX8953b+/eH2Qm6qPlTIL1Nc1fmQRr/+6GGWfi52+FQkND0d7e7tdrzoSqJshNTU3Ys2cPnE4ndu3ahX379gW6JLqbdDo43X8OneKpGBIyu0EbEjL1OdUsNFSbdQczZqI+Ws4kWF/T/JVJsPbvbtDwOFEUBd1ey3FsNhsURfFrDaq5SU8NC7KJiIiIKLBSUlLw22+/oaurC6Ojo/jkk0/w1FNP+bUG1fzTQg0Lssn/qqurA10CERERqUhoaCiOHj2KnJwcOJ1OFBcXw2g0+rcGv17Nh56eHsTGxnp+jomJwfnz5yfsd+zYMRw7dgwA4HA40N/f77cag83Q0FCgS0BOTg4ATJmjY3QUulu3ZnxeGR1FqAafG2rIhMZiJuqj5UyC9TXNX5kEa//uBi2PEwDIzc1FbgDXjatmgjxd3nc5hoeHa/ZGDbVQe//sCxdCt3jxjI+T0VGEqfzvNhW1Z/JfxEzUR6uZBPNrml9u0gvi/t0NWh0naqCaNchqWJBN/ldTU4OamppAl0FERETkoZoJshoWZJP/VVRUoKKiItBlEBEREXmoZomFGhZkExERERGpZoIMBH5BNhERERGRapZYEBERERGpASfIREREREReOEEmIiIiIvLCCTIRERERkRediEigi5itBQsWYPEsfmE4/cvhcCA0VFX3af7nMRP1YSbqw0zUh5moj9ozuXXrFlwuV6DLmJKmJ8g0N48++ija29sDXQZ5YSbqw0zUh5moDzNRH2YyN1xiQURERETkhRNkIiIiIiIvnCD/h+3evTvQJdA4zER9mIn6MBP1YSbqw0zmhmuQiYiIiIi88BNkIiIiIiIvnCATEREREXnhBFlFuru7kZmZiTVr1sBoNOLw4cMAgMHBQWRnZ8NgMCA7OxtDQ0MAgF9++QXp6elYtGgR3n77bc95bt++jdTUVJjNZhiNRlRVVU15TavViiVLlmDLli1jtm/fvh2rVq3C2rVrUVxcDLvdPunxR48eRUJCAnQ6Ha5fv+7ZLiIoLS1FQkICEhMTceHChVn3JZC0mMmd9mtra0NoaChqa2tn1ZNAm69M3JxOJ5KSkib029vJkydhMBhgMBhw8uRJz/bKykrExsYiIiLCZ80dHR0wmUxISEhAaWkp3Cvbtm3bBovFAovFAr1eD4vFMtN2BJzW8hgZGcHmzZuxevVqGI1G7Nu3z/PYt99+i+TkZE2PDze15OKr3+NNNU7cDh06NOG9Riu0mMdU4+nq1avIzMxEUlISEhMT0djYOON+aIKQaly7dk06OjpERGR4eFgMBoN0dnZKeXm5HDx4UEREDh48KBUVFSIi0tfXJ62trfLqq6/KW2+95TmPy+WSP//8U0RERkdHJTU1VVpaWia95ldffSX19fWyefPmMdsbGhrE5XKJy+WS/Px8ee+99yY9/sKFC9LV1SVxcXHS398/5nir1Soul0taWlokNTV1ll0JLC1m4ms/h8MhmZmZsmnTJvnss89m2ZXAmq9M3A4dOiQFBQUT+u02MDAgK1askIGBARkcHJQVK1bI4OCgiIi0tLTItWvX5N577/VZc0pKirS0tIjL5RKr1SqNjY0T9nn55ZflwIED02+ESmgtj7///luam5tFROSff/6RjIwMTx5dXV1y8eJFKSws1Oz4cFNLLr76PZ6vcXL16lV58skn5eGHHx7zXqMVWsxjqvH03HPPed5XOjs7JS4ubuYN0QB+gqwi0dHRSE5OBgBERkbikUceQU9PD+rq6lBUVAQAKCoqwhdffAEAiIqKQkpKCsLCwsacR6fTef7FZ7fbYbfbodPpJr1mVlYWIiMjJ2zPzc2FTqeDTqdDamoqbDbbpMcnJSVBr9dP2F5XV4edO3dCp9MhLS0NN27cQG9v77T6oCZazMTXfkeOHEFeXh6ioqJm1ggVma9MAMBms6GhoQG7du2a8npnz55FdnY2li5digceeADZ2dloamoCAKSlpSE6Otpnvb29vRgeHkZaWhp0Oh127tzpqc1NRPDpp5+ioKBgum1QDa3lER4ejszMTADAwoULkZyc7Bkjer0eiYmJWLBA+2+NasnFV7+93WmcvPTSS6iurp7ydVPttJYHMPV40ul0GB4eBgDcvHkTDz300PQboSHafxUIUleuXMEPP/yA9evXo6+vz/MkXb58Ofr6+u54vNPphMViQVRUFLKzs7F+/fpZ1WG32/HRRx/BarXO6Lienh7ExsZ6fo6JiUFPT8+salALrWUyfr+enh58/vnneP7552d1XTWaayZlZWWorq72OSGa63O5p6cHMTExPo8/d+4cli1bBoPBMO3zqpEW8vB248YNnD59GllZWbM6XivUkouvfvsaJ3V1dVAUBWaz+Y61aoEW8vBl//79qKmpQUxMDHJzc3HkyJEZHa8VnCCr0F9//YW8vDy8++67uO+++8Y85v5k8E5CQkLw448/wmazobW1FZcuXZpVLSUlJdi4cSM2bNgwq+ODhRYzGb9fWVkZ3nzzzaD4dAyYeyZnzpxBVFQU1q1bdzfLnJaPP/5Yk58ee9NaHg6HAwUFBSgtLUV8fLxfrhkIaslltv0eGRnBG2+8gddff31O11cLrecB/Pt69cwzz8Bms6GxsRGFhYVwuVxzqkeNguOdMojY7Xbk5eVh+/bt2Lp1KwBg2bJlnuUJvb29M/p6fMmSJcjMzERTUxPOnz/vuSGovr7+jsceOHAA/f39eOeddzzbcnJyYLFYfH61AwCKoqC7u9vzs81mg6Io065bTbSYyWT7tbe3Iz8/H3q9HrW1tSgpKZnwVb9WzEcm3333Herr66HX65Gfn4/m5mbs2LFjQiYzfS67vymwWCx47bXXoCjKmK8wxx/vcDhw6tQpbNu2bVa9UAMt5eG2e/duGAwGlJWVzeFvrm5qymV8v6c7Ti5fvoyuri6YzWbo9XrYbDYkJyfj999/n682+Y2W8vDlxIkTePrppwEA6enpuH37tiZvnLyjQC+Cpv9zuVxSWFgoe/bsGbN97969Yxbxl5eXj3m8qqpqzCL+P/74Q4aGhkREZGRkRDIyMuT06dNTXvfrr7+esND/+PHjkp6eLiMjI9OqffxNemfOnBlzk15KSsq0zqM2WsxkOvsVFRVp9iak+crE22T9dhsYGBC9Xi+Dg4MyODgoer1eBgYGxuwz05v0GhoaPI99+eWXsnHjRp/Hq5kW86isrJStW7eK0+mc9HEtjw83NeVyp367+RonbuPfa7RCi3m4jR9PVqtVPvzwQxER+fnnnyU6OlpcLte0zqUlnCCryLlz5wSAmEwmMZvNYjabpaGhQa5fvy5PPPGEJCQkSFZWludJ3tvbK4qiSGRkpNx///2iKIrcvHlTLl68KBaLRUwmkxiNRp93xmdkZMiDDz4o99xzjyiKIk1NTSIiEhISIvHx8Z46pjrH4cOHRVEUCQkJkejoaHn22WdF5N8Xg5KSEomPj5e1a9dKW1vbPHfLP7SYyXT20/IEYL4y8ebrjUZE5MSJE7Jy5UpZuXKlfPDBB57t5eXloiiK6HQ6URRFqqqqJj2+ra1NjEajxMfHywsvvDDmzaSoqEjef//9OXQksLSWR3d3twCQ1atXe+o9fvy4iIi0traKoigSHh4uS5culTVr1sxDhwJDLbn46vd4vsaJm1YnyFrMY6rx1NnZKY899pgkJiaK2WyWs2fPzkOH1If/1TQRERERkReuQSYiIiIi8sIJMhERERGRF06QiYiIiIi8cIJMREREROSFE2QiIiIiIi+cIBMR+cHAwIDnF/EvX74ciqLAYrEgIiICJSUlgS6PiIi88Ne8ERH52f79+xEREYG9e/cGuhQiIpoEP0EmIgqgb775Blu2bAHw78S5qKgIGzZsQFxcHE6dOoWKigqYTCZYrVbY7XYAQEdHBx5//HGsW7cOOTk5nv+qloiI5gcnyEREKnL58mU0Nzejvr4eO3bsQGZmJn766ScsXrwYDQ0NsNvtePHFF1FbW4uOjg4UFxejsrIy0GUTEQWV0EAXQERE/7dp0yaEhYXBZDLB6XTCarUCAEwmE65cuYJff/0Vly5dQnZ2NgDA6XQiOjo6kCUTEQUdTpCJiFRk0aJFAIAFCxYgLCwMOp3O87PD4YCIwGg0oqWlJZBlEhEFNS6xICLSkFWrVqG/v98zQbbb7ejs7AxwVUREwYUTZCIiDVm4cCFqa2vxyiuvwGw2w2Kx4Pvvvw90WUREQYW/5o2IiIiIyAs/QSYiIiIi8sIJMhERERGRF06QiYiIiIi8cIJMREREROSFE2QiIiIiIi+cIBMREREReeEEmYiIiIjIy/8AXFQihURk+hAAAAAASUVORK5CYII=\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "WindStats\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "ProphetDetector\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "for model in [model1, model2, model3]:\n", " print(type(model).__name__)\n", " fig, ax = model.plot_anomaly(\n", " time_series=test_data, time_series_prev=train_data,\n", " filter_scores=True, plot_time_series_prev=True)\n", " plot_anoms(ax=ax, anomaly_labels=test_labels)\n", " plt.show()\n", " print()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "So all the individual models generate quite a few false positives. Let's see how the ensemble does:" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, ax = ensemble.plot_anomaly(\n", " time_series=test_data, time_series_prev=train_data,\n", " filter_scores=True, plot_time_series_prev=True)\n", "plot_anoms(ax=ax, anomaly_labels=test_labels)\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "So the ensemble misses one of the three anomalies in the test split, but it also greatly reduces the number of false positives relative to the other models." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Saving & Loading Models\n", "\n", "All models have a `save()` method and `load()` class method. Models may also be loaded with the assistance of the `ModelFactory`, which works for arbitrary models. The `save()` method creates a new directory at the specified path, where it saves a `json` file representing the model's config, as well as a binary file for the model's state.\n", "\n", "We will demonstrate these behaviors using our `IsolationForest` model (`model1`) for concreteness. Note that the config explicitly tracks the transform (to pre-process the data), the calibrator (to transform raw anomaly scores into z-scores), the thresholding rule (to sparsify the calibrated anomaly scores)." ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "IsolationForest Config\n", "{'calibrator': {'abs_score': True,\n", " 'anchors': [[0.38992633996347176, 0.0],\n", " [0.41877507813617143, 0.5],\n", " [0.445336977389891, 1.0],\n", " [0.47974261897360404, 1.5],\n", " [0.5271631189090943, 2.0],\n", " [0.8301789920204418, 4.032894437734716],\n", " [1.0, 5.032894437734716]],\n", " 'max_score': 1.0,\n", " 'name': 'AnomScoreCalibrator'},\n", " 'enable_calibrator': True,\n", " 'enable_threshold': True,\n", " 'max_n_samples': 1.0,\n", " 'model_path': '/Users/abhatnagar/Desktop/Merlion/examples/anomaly/models/isf/model.pkl',\n", " 'n_estimators': 100,\n", " 'threshold': {'abs_score': True,\n", " 'alm_suppress_minutes': 120,\n", " 'alm_threshold': 3.2263155501877727,\n", " 'alm_window_minutes': 60,\n", " 'min_alm_in_window': 2,\n", " 'name': 'AggregateAlarms'},\n", " 'transform': {'name': 'TransformSequence',\n", " 'transforms': [{'name': 'DifferenceTransform'},\n", " {'multivar_skip': True,\n", " 'name': 'Shingle',\n", " 'size': 2,\n", " 'stride': 1}]}}\n" ] } ], "source": [ "import json\n", "import os\n", "import pprint\n", "from merlion.models.factory import ModelFactory\n", "\n", "# Save the model\n", "os.makedirs(\"models\", exist_ok=True)\n", "path = os.path.join(\"models\", \"isf\")\n", "model1.save(path)\n", "\n", "# Print the config saved\n", "pp = pprint.PrettyPrinter()\n", "with open(os.path.join(path, \"config.json\")) as f:\n", " print(f\"{type(model1).__name__} Config\")\n", " pp.pprint(json.load(f))\n", "\n", "# Load the model using Prophet.load()\n", "model2_loaded = IsolationForest.load(dirname=path)\n", "\n", "# Load the model using the ModelFactory\n", "model2_factory_loaded = ModelFactory.load(name=\"IsolationForest\", model_path=path)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can do the same exact thing with ensembles! Note that the ensemble saves each of its sub-models in a different sub-directory, which it tracks manually." ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Ensemble Config\n", "{'calibrator': {'abs_score': True,\n", " 'anchors': None,\n", " 'max_score': 1000,\n", " 'name': 'AnomScoreCalibrator'},\n", " 'combiner': {'abs_score': True, 'n_models': 3, 'name': 'Mean'},\n", " 'enable_calibrator': False,\n", " 'enable_threshold': True,\n", " 'model_paths': [['IsolationForest',\n", " '/Users/abhatnagar/Desktop/Merlion/examples/anomaly/models/ensemble/0'],\n", " ['WindStats',\n", " '/Users/abhatnagar/Desktop/Merlion/examples/anomaly/models/ensemble/1'],\n", " ['ProphetDetector',\n", " '/Users/abhatnagar/Desktop/Merlion/examples/anomaly/models/ensemble/2']],\n", " 'threshold': {'abs_score': True,\n", " 'alm_suppress_minutes': 120,\n", " 'alm_threshold': 4,\n", " 'alm_window_minutes': 60,\n", " 'min_alm_in_window': 2,\n", " 'name': 'AggregateAlarms'},\n", " 'transform': {'name': 'Identity'}}\n" ] } ], "source": [ "# Save the ensemble\n", "path = os.path.join(\"models\", \"ensemble\")\n", "ensemble.save(path)\n", "\n", "# Print the config saved. Note that we've saved all individual models,\n", "# and their paths are specified under the model_paths key.\n", "pp = pprint.PrettyPrinter()\n", "with open(os.path.join(path, \"config.json\")) as f:\n", " print(f\"Ensemble Config\")\n", " pp.pprint(json.load(f))\n", "\n", "# Load the selector\n", "selector_loaded = DetectorEnsemble.load(dirname=path)\n", "\n", "# Load the selector using the ModelFactory\n", "selector_factory_loaded = ModelFactory.load(name=\"DetectorEnsemble\", model_path=path)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Simulating Live Model Deployment\n", "\n", "A typical model deployment scenario is as follows:\n", "1. Train an initial model on some recent historical data, optionally with labels.\n", "1. At a regular interval `retrain_freq` (e.g. once per week), retrain the entire model unsupervised (i.e. with no labels) on the most recent data.\n", "1. Obtain the model's predicted anomaly scores for the time series values that occur between re-trainings. We perform this operation in batch, but a deployment scenario may do this in streaming.\n", "1. Optionally, specify a maximum amount of data (`train_window`) that the model should use for training (e.g. the most recent 2 weeks of data).\n", "\n", "We provide a `TSADEvaluator` object which simulates the above deployment scenario, and also allows a user to evaluate the quality of the forecaster according to an evaluation metric of their choice. We illustrate an example below using the ensemble." ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [], "source": [ "# Initialize the evaluator\n", "from merlion.evaluate.anomaly import TSADEvaluator, TSADEvaluatorConfig\n", "\n", "evaluator = TSADEvaluator(model=ensemble, config=TSADEvaluatorConfig(retrain_freq=\"7d\"))" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "INFO:merlion.models.forecast.prophet:Add seasonality 15\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "INFO:fbprophet:Disabling weekly seasonality. Run prophet with weekly_seasonality=True to override this.\n", "TSADEvaluator: 10%|█ | 604800/5783700 [00:00<00:04, 1039930.97it/s]INFO:merlion.models.forecast.prophet:Add seasonality 13\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 21%|██ | 1209600/5783700 [00:03<00:15, 291306.22it/s]INFO:merlion.models.forecast.prophet:Add seasonality 13\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 31%|███▏ | 1814400/5783700 [00:07<00:19, 204205.27it/s]INFO:merlion.models.forecast.prophet:Add seasonality 15\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 42%|████▏ | 2419200/5783700 [00:10<00:16, 200797.88it/s]INFO:merlion.models.forecast.prophet:Add seasonality 15\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 52%|█████▏ | 3024000/5783700 [00:16<00:18, 150898.03it/s]INFO:merlion.models.forecast.prophet:Add seasonality 15\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 63%|██████▎ | 3628800/5783700 [00:20<00:14, 151825.96it/s]INFO:merlion.models.forecast.prophet:Add seasonality 12\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 73%|███████▎ | 4233600/5783700 [00:26<00:12, 127252.90it/s]INFO:merlion.models.forecast.prophet:Add seasonality 12\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 84%|████████▎ | 4838400/5783700 [00:31<00:07, 128260.14it/s]INFO:merlion.models.forecast.prophet:Add seasonality 12\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 94%|█████████▍| 5443200/5783700 [00:40<00:03, 99713.90it/s] INFO:merlion.models.forecast.prophet:Add seasonality 12\n", "INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.\n", "TSADEvaluator: 100%|██████████| 5783700/5783700 [00:49<00:00, 116554.57it/s]\n" ] } ], "source": [ "# The kwargs we would provide to ensemble.train() for the initial training\n", "# Note that we are training the ensemble unsupervised.\n", "train_kwargs = {\"anomaly_labels\": None}\n", "\n", "# We will use the default kwargs for re-training (these leave the\n", "# post-rules unchanged, since there are no new labels)\n", "retrain_kwargs = None\n", "\n", "# We call evaluator.get_predict() to get the time series of anomaly scores\n", "# produced by the anomaly detector when deployed in this manner\n", "train_scores, test_scores = evaluator.get_predict(\n", " train_vals=train_data, test_vals=test_data,\n", " train_kwargs=train_kwargs, retrain_kwargs=retrain_kwargs\n", ")" ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Ensemble Performance\n", "Precision: 0.6667\n", "Recall: 0.6667\n", "F1: 0.6667\n", "MTTD: 1 days 10:27:30\n", "\n" ] } ], "source": [ "# Now let's evaluate how we did.\n", "precision = evaluator.evaluate(ground_truth=test_labels, predict=test_scores, metric=TSADMetric.Precision)\n", "recall = evaluator.evaluate(ground_truth=test_labels, predict=test_scores, metric=TSADMetric.Recall)\n", "f1 = evaluator.evaluate(ground_truth=test_labels, predict=test_scores, metric=TSADMetric.F1)\n", "mttd = evaluator.evaluate(ground_truth=test_labels, predict=test_scores, metric=TSADMetric.MeanTimeToDetect)\n", "print(\"Ensemble Performance\")\n", "print(f\"Precision: {precision:.4f}\")\n", "print(f\"Recall: {recall:.4f}\")\n", "print(f\"F1: {f1:.4f}\")\n", "print(f\"MTTD: {mttd}\")\n", "print()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In this case, we see that by simply re-training the ensemble weekly in an unsupervised manner, we have increased the precision from $\\frac{2}{5}$ to $\\frac{2}{3}$, while leaving unchanged the recall and mean time to detect. This is due to data drift over time." ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.5" } }, "nbformat": 4, "nbformat_minor": 4 }