From 0c0f39a4a268c184e0035b48285ab2146f3eb4e3 Mon Sep 17 00:00:00 2001 From: moocarme Date: Fri, 10 Jan 2020 23:53:39 -0500 Subject: [PATCH] cleanup --- Chapter01/Exercise1.01/Exercise1_01.ipynb | 2 +- Chapter01/Exercise1.02/Exercise1_02.ipynb | 38 ++-- Chapter04/Activity4.02/Activity4_02.ipynb | 0 Chapter04/Activity4.03/Activity4_03.ipynb | 0 Chapter04/Exercise4.02/Exercise4_02.ipynb | 0 Chapter04/Exercise4.03/Exercise4_03.ipynb | 0 Chapter05/Activity5.01/Activity5_01.ipynb | 0 Chapter05/Activity5.03/Activity5_03.ipynb | 0 Chapter05/Exercise5.01/Exercise5_01.ipynb | 0 Chapter05/Exercise5.02/Exercise5_02.ipynb | 0 Chapter06/Activity6.01/Activity6_01.ipynb | 14 -- Chapter06/Activity6.02/Activity6_02.ipynb | 2 +- Chapter06/Exercise6.03/Exercise6_03.ipynb | 10 +- Chapter07/Activity7.01/Activity7_01.ipynb | 0 Chapter07/Activity7.02/Activity7_02.ipynb | 0 Chapter07/Exercise7.01/Exercise7_01.ipynb | 0 Chapter07/Exercise7.02/Exercise7_02.ipynb | 0 Chapter07/Exercise7.03/Exercise7_03.ipynb | 0 Chapter07/Exercise7.04/Exercise7_04.ipynb | 0 Chapter08/Activity8.01/Activity8_01.ipynb | 0 Chapter08/Activity8.02/Activity8_02.ipynb | 0 Chapter08/Exercise8.01/Exercise8_01.ipynb | 0 Chapter08/Exercise8.02/Exercise8_02.ipynb | 0 Chapter08/Exercise8.03/Exercise8_03.ipynb | 0 Chapter08/Exercise8.04/Exercise8_04.ipynb | 0 Chapter09/Activity9.01/Activity9_01.ipynb | 11 +- Chapter09/Activity9.02/Activity9_02.ipynb | 11 +- Chapter09/Activity9.03/Activity9_03.ipynb | 11 +- Chapter09/Exercise9.01/Exercise9_01.ipynb | 222 +++++++++++----------- Chapter09/Exercise9.02/Exercise9_02.ipynb | 11 +- 30 files changed, 166 insertions(+), 166 deletions(-) mode change 100644 => 100755 Chapter04/Activity4.02/Activity4_02.ipynb mode change 100644 => 100755 Chapter04/Activity4.03/Activity4_03.ipynb mode change 100644 => 100755 Chapter04/Exercise4.02/Exercise4_02.ipynb mode change 100644 => 100755 Chapter04/Exercise4.03/Exercise4_03.ipynb mode change 100644 => 100755 Chapter05/Activity5.01/Activity5_01.ipynb mode change 100755 => 100644 Chapter05/Activity5.03/Activity5_03.ipynb mode change 100644 => 100755 Chapter05/Exercise5.01/Exercise5_01.ipynb mode change 100644 => 100755 Chapter05/Exercise5.02/Exercise5_02.ipynb mode change 100644 => 100755 Chapter07/Activity7.01/Activity7_01.ipynb mode change 100644 => 100755 Chapter07/Activity7.02/Activity7_02.ipynb mode change 100644 => 100755 Chapter07/Exercise7.01/Exercise7_01.ipynb mode change 100644 => 100755 Chapter07/Exercise7.02/Exercise7_02.ipynb mode change 100644 => 100755 Chapter07/Exercise7.03/Exercise7_03.ipynb mode change 100644 => 100755 Chapter07/Exercise7.04/Exercise7_04.ipynb mode change 100644 => 100755 Chapter08/Activity8.01/Activity8_01.ipynb mode change 100644 => 100755 Chapter08/Activity8.02/Activity8_02.ipynb mode change 100644 => 100755 Chapter08/Exercise8.01/Exercise8_01.ipynb mode change 100644 => 100755 Chapter08/Exercise8.02/Exercise8_02.ipynb mode change 100644 => 100755 Chapter08/Exercise8.03/Exercise8_03.ipynb mode change 100644 => 100755 Chapter08/Exercise8.04/Exercise8_04.ipynb mode change 100644 => 100755 Chapter09/Activity9.01/Activity9_01.ipynb mode change 100644 => 100755 Chapter09/Activity9.02/Activity9_02.ipynb mode change 100644 => 100755 Chapter09/Activity9.03/Activity9_03.ipynb mode change 100644 => 100755 Chapter09/Exercise9.01/Exercise9_01.ipynb mode change 100644 => 100755 Chapter09/Exercise9.02/Exercise9_02.ipynb diff --git a/Chapter01/Exercise1.01/Exercise1_01.ipynb b/Chapter01/Exercise1.01/Exercise1_01.ipynb index 5a33e505..be718b44 100644 --- a/Chapter01/Exercise1.01/Exercise1_01.ipynb +++ b/Chapter01/Exercise1.01/Exercise1_01.ipynb @@ -776,7 +776,7 @@ "outputs": [], "source": [ "feats.to_csv('../data/OSI_feats.csv', index=False)\n", - "target.to_csv('../data/OSI_target.csv', header='Class', index=False)" + "target.to_csv('../data/OSI_target.csv', header='Revenue', index=False)" ] } ], diff --git a/Chapter01/Exercise1.02/Exercise1_02.ipynb b/Chapter01/Exercise1.02/Exercise1_02.ipynb index e52a1ba1..c6a70283 100755 --- a/Chapter01/Exercise1.02/Exercise1_02.ipynb +++ b/Chapter01/Exercise1.02/Exercise1_02.ipynb @@ -859,22 +859,22 @@ { "data": { "text/plain": [ - "array([[,\n", - " ,\n", - " ,\n", - " ],\n", - " [,\n", - " ,\n", - " ,\n", - " ],\n", - " [,\n", - " ,\n", - " ,\n", - " ],\n", - " [,\n", - " ,\n", - " ,\n", - " ]],\n", + "array([[,\n", + " ,\n", + " ,\n", + " ],\n", + " [,\n", + " ,\n", + " ,\n", + " ],\n", + " [,\n", + " ,\n", + " ,\n", + " ],\n", + " [,\n", + " ,\n", + " ,\n", + " ]],\n", " dtype=object)" ] }, @@ -940,7 +940,7 @@ { "data": { "text/plain": [ - "" + "" ] }, "execution_count": 6, @@ -1104,7 +1104,7 @@ { "data": { "text/plain": [ - "" + "" ] }, "execution_count": 10, @@ -1850,7 +1850,7 @@ { "data": { "text/plain": [ - "" + "" ] }, "execution_count": 25, diff --git a/Chapter04/Activity4.02/Activity4_02.ipynb b/Chapter04/Activity4.02/Activity4_02.ipynb old mode 100644 new mode 100755 diff --git a/Chapter04/Activity4.03/Activity4_03.ipynb b/Chapter04/Activity4.03/Activity4_03.ipynb old mode 100644 new mode 100755 diff --git a/Chapter04/Exercise4.02/Exercise4_02.ipynb b/Chapter04/Exercise4.02/Exercise4_02.ipynb old mode 100644 new mode 100755 diff --git a/Chapter04/Exercise4.03/Exercise4_03.ipynb b/Chapter04/Exercise4.03/Exercise4_03.ipynb old mode 100644 new mode 100755 diff --git a/Chapter05/Activity5.01/Activity5_01.ipynb b/Chapter05/Activity5.01/Activity5_01.ipynb old mode 100644 new mode 100755 diff --git a/Chapter05/Activity5.03/Activity5_03.ipynb b/Chapter05/Activity5.03/Activity5_03.ipynb old mode 100755 new mode 100644 diff --git a/Chapter05/Exercise5.01/Exercise5_01.ipynb b/Chapter05/Exercise5.01/Exercise5_01.ipynb old mode 100644 new mode 100755 diff --git a/Chapter05/Exercise5.02/Exercise5_02.ipynb b/Chapter05/Exercise5.02/Exercise5_02.ipynb old mode 100644 new mode 100755 diff --git a/Chapter06/Activity6.01/Activity6_01.ipynb b/Chapter06/Activity6.01/Activity6_01.ipynb index ae623639..221a5254 100644 --- a/Chapter06/Activity6.01/Activity6_01.ipynb +++ b/Chapter06/Activity6.01/Activity6_01.ipynb @@ -257,20 +257,6 @@ "X_test = pd.DataFrame(X_test, columns = X_train.columns)" ] }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "#Convert DataFrame to Numpy array\n", - "\n", - "#x_train=xtrain.values\n", - "#x_test=xtest.values\n", - "#y_train=ytrain.values\n", - "#y_test=ytest.values\n" - ] - }, { "cell_type": "markdown", "metadata": {}, diff --git a/Chapter06/Activity6.02/Activity6_02.ipynb b/Chapter06/Activity6.02/Activity6_02.ipynb index deb8de95..a128c680 100644 --- a/Chapter06/Activity6.02/Activity6_02.ipynb +++ b/Chapter06/Activity6.02/Activity6_02.ipynb @@ -321,7 +321,7 @@ ], "source": [ "# Fit the Model\n", - "model.fit(X_train, y_train, epochs=100, batch_size=20, verbose=1, validation_split=0.2)" + "model.fit(X_train, y_train, epochs=100, batch_size=20, verbose=1, validation_split=0.2, shuffle=False)" ] }, { diff --git a/Chapter06/Exercise6.03/Exercise6_03.ipynb b/Chapter06/Exercise6.03/Exercise6_03.ipynb index fd6c7517..c7dcfa95 100644 --- a/Chapter06/Exercise6.03/Exercise6_03.ipynb +++ b/Chapter06/Exercise6.03/Exercise6_03.ipynb @@ -214,13 +214,7 @@ "38400/38400 [==============================] - 6s 150us/step - loss: 0.0221 - accuracy: 0.9945 - val_loss: 0.0299 - val_accuracy: 0.9925\n", "Epoch 54/100\n", "38400/38400 [==============================] - 5s 135us/step - loss: 0.0237 - accuracy: 0.9940 - val_loss: 0.0472 - val_accuracy: 0.9925\n", - "Epoch 55/100\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ + "Epoch 55/100\n", "38400/38400 [==============================] - 6s 152us/step - loss: 0.0214 - accuracy: 0.9939 - val_loss: 0.0414 - val_accuracy: 0.9925\n", "Epoch 56/100\n", "38400/38400 [==============================] - 5s 138us/step - loss: 0.0224 - accuracy: 0.9936 - val_loss: 0.0426 - val_accuracy: 0.9933\n", @@ -327,7 +321,7 @@ ], "source": [ "# Fit the Model\n", - "model.fit(X_train, y_train, epochs=100, batch_size=20, verbose=1, validation_split=0.2)" + "model.fit(X_train, y_train, epochs=100, batch_size=20, verbose=1, validation_split=0.2, shuffle=False)" ] }, { diff --git a/Chapter07/Activity7.01/Activity7_01.ipynb b/Chapter07/Activity7.01/Activity7_01.ipynb old mode 100644 new mode 100755 diff --git a/Chapter07/Activity7.02/Activity7_02.ipynb b/Chapter07/Activity7.02/Activity7_02.ipynb old mode 100644 new mode 100755 diff --git a/Chapter07/Exercise7.01/Exercise7_01.ipynb b/Chapter07/Exercise7.01/Exercise7_01.ipynb old mode 100644 new mode 100755 diff --git a/Chapter07/Exercise7.02/Exercise7_02.ipynb b/Chapter07/Exercise7.02/Exercise7_02.ipynb old mode 100644 new mode 100755 diff --git a/Chapter07/Exercise7.03/Exercise7_03.ipynb b/Chapter07/Exercise7.03/Exercise7_03.ipynb old mode 100644 new mode 100755 diff --git a/Chapter07/Exercise7.04/Exercise7_04.ipynb b/Chapter07/Exercise7.04/Exercise7_04.ipynb old mode 100644 new mode 100755 diff --git a/Chapter08/Activity8.01/Activity8_01.ipynb b/Chapter08/Activity8.01/Activity8_01.ipynb old mode 100644 new mode 100755 diff --git a/Chapter08/Activity8.02/Activity8_02.ipynb b/Chapter08/Activity8.02/Activity8_02.ipynb old mode 100644 new mode 100755 diff --git a/Chapter08/Exercise8.01/Exercise8_01.ipynb b/Chapter08/Exercise8.01/Exercise8_01.ipynb old mode 100644 new mode 100755 diff --git a/Chapter08/Exercise8.02/Exercise8_02.ipynb b/Chapter08/Exercise8.02/Exercise8_02.ipynb old mode 100644 new mode 100755 diff --git a/Chapter08/Exercise8.03/Exercise8_03.ipynb b/Chapter08/Exercise8.03/Exercise8_03.ipynb old mode 100644 new mode 100755 diff --git a/Chapter08/Exercise8.04/Exercise8_04.ipynb b/Chapter08/Exercise8.04/Exercise8_04.ipynb old mode 100644 new mode 100755 diff --git a/Chapter09/Activity9.01/Activity9_01.ipynb b/Chapter09/Activity9.01/Activity9_01.ipynb old mode 100644 new mode 100755 index 11d69ddd..3fe290a6 --- a/Chapter09/Activity9.01/Activity9_01.ipynb +++ b/Chapter09/Activity9.01/Activity9_01.ipynb @@ -16,7 +16,8 @@ "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", - "import pandas as pd" + "import pandas as pd\n", + "from tensorflow import random" ] }, { @@ -131,7 +132,7 @@ } ], "source": [ - "dataset_training = pd.read_csv('AMZN_train.csv')\n", + "dataset_training = pd.read_csv('../AMZN_train.csv')\n", "dataset_training.head()" ] }, @@ -315,7 +316,9 @@ "metadata": {}, "outputs": [], "source": [ - "np.random.seed(1)\n", + "seed = 1\n", + "np.random.seed(seed)\n", + "random.set_seed(seed)\n", "\n", "# initialize model\n", "model = Sequential()" @@ -616,7 +619,7 @@ } ], "source": [ - "dataset_testing = pd.read_csv('AMZN_test.csv')\n", + "dataset_testing = pd.read_csv('../AMZN_test.csv')\n", "actual_stock_price = dataset_testing[['Open']].values\n", "actual_stock_price" ] diff --git a/Chapter09/Activity9.02/Activity9_02.ipynb b/Chapter09/Activity9.02/Activity9_02.ipynb old mode 100644 new mode 100755 index d83e279e..e9f834ec --- a/Chapter09/Activity9.02/Activity9_02.ipynb +++ b/Chapter09/Activity9.02/Activity9_02.ipynb @@ -16,7 +16,8 @@ "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", - "import pandas as pd" + "import pandas as pd\n", + "from tensorflow import random" ] }, { @@ -131,7 +132,7 @@ } ], "source": [ - "dataset_training = pd.read_csv('AMZN_train.csv')\n", + "dataset_training = pd.read_csv('../AMZN_train.csv')\n", "dataset_training.head()" ] }, @@ -255,7 +256,9 @@ "metadata": {}, "outputs": [], "source": [ - "np.random.seed(1)\n", + "seed = 1\n", + "np.random.seed(seed)\n", + "random.set_seed(seed)\n", "\n", "model = Sequential()" ] @@ -559,7 +562,7 @@ } ], "source": [ - "dataset_testing = pd.read_csv('AMZN_test.csv')\n", + "dataset_testing = pd.read_csv('../AMZN_test.csv')\n", "actual_stock_price = dataset_testing[['Open']].values\n", "actual_stock_price" ] diff --git a/Chapter09/Activity9.03/Activity9_03.ipynb b/Chapter09/Activity9.03/Activity9_03.ipynb old mode 100644 new mode 100755 index 5dcb0be5..a9768233 --- a/Chapter09/Activity9.03/Activity9_03.ipynb +++ b/Chapter09/Activity9.03/Activity9_03.ipynb @@ -16,7 +16,8 @@ "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", - "import pandas as pd" + "import pandas as pd\n", + "from tensorflow import random" ] }, { @@ -131,7 +132,7 @@ } ], "source": [ - "dataset_training = pd.read_csv('AMZN_train.csv')\n", + "dataset_training = pd.read_csv('../AMZN_train.csv')\n", "dataset_training.head()" ] }, @@ -315,7 +316,9 @@ "metadata": {}, "outputs": [], "source": [ - "np.random.seed(1)\n", + "seed = 1\n", + "np.random.seed(seed)\n", + "random.set_seed(seed)\n", "\n", "model = Sequential()" ] @@ -615,7 +618,7 @@ } ], "source": [ - "dataset_testing = pd.read_csv('AMZN_test.csv')\n", + "dataset_testing = pd.read_csv('../AMZN_test.csv')\n", "actual_stock_price = dataset_testing[['Open']].values\n", "actual_stock_price" ] diff --git a/Chapter09/Exercise9.01/Exercise9_01.ipynb b/Chapter09/Exercise9.01/Exercise9_01.ipynb old mode 100644 new mode 100755 index 04ca4089..89cc6f08 --- a/Chapter09/Exercise9.01/Exercise9_01.ipynb +++ b/Chapter09/Exercise9.01/Exercise9_01.ipynb @@ -16,7 +16,8 @@ "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", - "import pandas as pd" + "import pandas as pd\n", + "from tensorflow import random" ] }, { @@ -131,7 +132,7 @@ } ], "source": [ - "dataset_training = pd.read_csv('GOOG_train.csv')\n", + "dataset_training = pd.read_csv('../GOOG_train.csv')\n", "dataset_training.head()" ] }, @@ -322,8 +323,9 @@ "metadata": {}, "outputs": [], "source": [ - "np.random.seed(1)\n", - "\n", + "seed = 1\n", + "np.random.seed(seed)\n", + "random.set_seed(seed)\n", "# initialize model\n", "model = Sequential()" ] @@ -373,211 +375,217 @@ "output_type": "stream", "text": [ "Epoch 1/100\n", - "1198/1198 [==============================] - 5s 4ms/step - loss: 0.0232\n", + "1198/1198 [==============================] - 5s 4ms/step - loss: 0.0240\n", "Epoch 2/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0022\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0023\n", "Epoch 3/100\n", "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0023\n", "Epoch 4/100\n", - "1198/1198 [==============================] - 4s 4ms/step - loss: 0.0019\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0021\n", "Epoch 5/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0018\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0019\n", "Epoch 6/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0021\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0022\n", "Epoch 7/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0022\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0023\n", "Epoch 8/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0021\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0022\n", "Epoch 9/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0019\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0020\n", "Epoch 10/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0017\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0018\n", "Epoch 11/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0017\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0018\n", "Epoch 12/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0016\n", - "Epoch 13/100\n", "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0017\n", + "Epoch 13/100\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0018\n", "Epoch 14/100\n", - "1198/1198 [==============================] - 4s 4ms/step - loss: 0.0014\n", - "Epoch 15/100\n", "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0015\n", + "Epoch 15/100\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0016\n", "Epoch 16/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0017\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0017\n", "Epoch 17/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0013\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0014\n", "Epoch 18/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0012\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0013\n", "Epoch 19/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0014\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0015\n", "Epoch 20/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0013\n", - "Epoch 21/100\n", "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0014\n", + "Epoch 21/100\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0016\n", "Epoch 22/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0012\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0013\n", "Epoch 23/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0010\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0011\n", "Epoch 24/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 9.7084e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 0.0011\n", "Epoch 25/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 9.8500e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0011\n", "Epoch 26/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 8.8464e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0010\n", "Epoch 27/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 8.1715e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 9.2147e-04\n", "Epoch 28/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0010\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 0.0011\n", "Epoch 29/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 9.1452e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 9.7054e-04\n", "Epoch 30/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 7.6600e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 8.5618e-04\n", "Epoch 31/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 6.9206e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 7.8764e-04\n", "Epoch 32/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 6.1541e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 6.8009e-04\n", "Epoch 33/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 8.5503e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 8.8178e-04\n", "Epoch 34/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 6.2782e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 8.0605e-04\n", "Epoch 35/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 6.8570e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 7.5420e-04\n", "Epoch 36/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 6.1862e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 6.4858e-04\n", "Epoch 37/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 5.4806e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.9398e-04\n", "Epoch 38/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 5.3405e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.9324e-04\n", "Epoch 39/100\n", - "1198/1198 [==============================] - 4s 4ms/step - loss: 6.2223e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 6.9292e-04\n", "Epoch 40/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.9710e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.5480e-04\n", "Epoch 41/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.7666e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.2705e-04\n", "Epoch 42/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.8689e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.4885e-04\n", "Epoch 43/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 5.7784e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 6.4513e-04\n", "Epoch 44/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 5.9633e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 6.8525e-04\n", "Epoch 45/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 5.2600e-04\n", + "1198/1198 [==============================] - 4s 4ms/step - loss: 6.2586e-04\n", "Epoch 46/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 6.3314e-04\n", + "1198/1198 [==============================] - 4s 4ms/step - loss: 6.6320e-04\n", "Epoch 47/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.9926e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 5.3459e-04\n", "Epoch 48/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.8287e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 5.1322e-04\n", "Epoch 49/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 5.4634e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.5603e-04\n", "Epoch 50/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 5.7624e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.8515e-04\n", "Epoch 51/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.4091e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.5264e-04\n", "Epoch 52/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.8870e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.1439e-04\n", "Epoch 53/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.4009e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.5409e-04\n", "Epoch 54/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.2719e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.4053e-04\n", "Epoch 55/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.1622e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.3687e-04\n", "Epoch 56/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.4584e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.6289e-04\n", "Epoch 57/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.0149e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.0678e-04\n", "Epoch 58/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.2947e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.4138e-04\n", "Epoch 59/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.9694e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.2136e-04\n", "Epoch 60/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.7688e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.9160e-04\n", "Epoch 61/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.0414e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.1005e-04\n", "Epoch 62/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.5619e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.6545e-04\n", "Epoch 63/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.7291e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.7596e-04\n", "Epoch 64/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.6891e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.8530e-04\n", "Epoch 65/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.1369e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.1432e-04\n", "Epoch 66/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.8439e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.9502e-04\n", "Epoch 67/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.2193e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.2827e-04\n", "Epoch 68/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.7382e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.8653e-04\n", "Epoch 69/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.8548e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.8565e-04\n", "Epoch 70/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.6493e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.4341e-04\n", "Epoch 71/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.9909e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.1007e-04\n", "Epoch 72/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.7788e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.8280e-04\n", "Epoch 73/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 5.2327e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 5.2328e-04\n", "Epoch 74/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.7966e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.3108e-04\n", "Epoch 75/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.4527e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.4717e-04\n", "Epoch 76/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.8860e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.8701e-04\n", "Epoch 77/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.6920e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.7963e-04\n", "Epoch 78/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.0639e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.0789e-04\n", "Epoch 79/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.6981e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.5674e-04\n", "Epoch 80/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.5855e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.6300e-04\n", "Epoch 81/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.2628e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.3991e-04\n", "Epoch 82/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.7443e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.8069e-04\n", "Epoch 83/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.4240e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.4161e-04\n", "Epoch 84/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.5120e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.6345e-04\n", "Epoch 85/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.1406e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.2626e-04\n", "Epoch 86/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.8843e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 5.7530e-04\n", "Epoch 87/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 3.9933e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.4177e-04\n", "Epoch 88/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.9938e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.8598e-04\n", "Epoch 89/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.5242e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.5695e-04\n", "Epoch 90/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.9862e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.0539e-04\n", "Epoch 91/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.5978e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.6065e-04\n", "Epoch 92/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.5843e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.6448e-04\n", "Epoch 93/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.0026e-04\n", - "Epoch 94/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 4.2905e-04\n", + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.0693e-04\n", + "Epoch 94/100\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1198/1198 [==============================] - 4s 3ms/step - loss: 4.2961e-04\n", "Epoch 95/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.4939e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.4710e-04\n", "Epoch 96/100\n", - "1198/1198 [==============================] - 3s 3ms/step - loss: 4.7370e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 4.9433e-04\n", "Epoch 97/100\n", - "1198/1198 [==============================] - 4s 4ms/step - loss: 3.9417e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.9799e-04\n", "Epoch 98/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.2308e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.3037e-04\n", "Epoch 99/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.8040e-04\n", + "1198/1198 [==============================] - 3s 3ms/step - loss: 3.9239e-04\n", "Epoch 100/100\n", - "1198/1198 [==============================] - 4s 3ms/step - loss: 3.7623e-04\n" + "1198/1198 [==============================] - 4s 3ms/step - loss: 3.7573e-04\n" ] }, { "data": { "text/plain": [ - "" + "" ] }, "execution_count": 10, @@ -637,7 +645,7 @@ } ], "source": [ - "dataset_testing = pd.read_csv(\"GOOG_test.csv\")\n", + "dataset_testing = pd.read_csv(\"../GOOG_test.csv\")\n", "actual_stock_price = dataset_testing[['Open']].values\n", "actual_stock_price" ] @@ -675,7 +683,7 @@ "outputs": [ { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOydd3hUxdeA30mBhN67SBUEpEsERREBG3wqAoqIiBXE9lNR7CCgKGBDwYY0BUWKSJOiBAQiNUASeichkEBICAnp5/tjNhhCyu7NbnYT5n2e++zuvXNmzm42e+7MmXOOEhEMBoPBYMgLL3crYDAYDAbPxxgLg8FgMOSLMRYGg8FgyBdjLAwGg8GQL8ZYGAwGgyFfjLEwGAwGQ74YY2Eodiil6imlRCnlY3u9XCk1qBDGHamU+snV4+SjQxelVLiL+u6slNrnir4Nno8xFga3oJQ6qpS6qJS6oJQ6rZSarpQq44qxRORuEZlhp07dXKGDrf+3lFJHbO85XCn1a5ZrgUqpp1w1di76PK6USrfpc14ptUMp1TO39iLyj4g0KUwdDZ6DMRYGd9JLRMoAbYH2wDvZGyhNkf+e2mY2A4FutvfcHvjLvVoBEGTTpwIwFZirlKqYvVHmLM1w9VLk/wkNRR8RiQCWAy3g0l32WKXUBiARaKCUKq+UmqqUilRKRSilxiilvG3tvZVSE5RSZ5RSh4F7s/af/a5dKfW0UmqPUipeKbVbKdVWKTULqAsstt1pv25re5NSaqNSKlYptVMp1SVLP/WVUmtt/awCquTxNm8EVojIIdt7PiUi39n6GQt0Br6yjf2V7XwnpdQWpVSc7bFTlrErKaWmKaVOKqXOKaV+z2lQpdSLtvdYJ5+/QQbwI+APNMxczlJKvaGUOgVMy77EpZS6Rim1QCkVrZQ6m6m37doTts/4nFJqhVLq2rzGNxQBRMQc5ij0AziKvssGuAYIA0bbXgcCx4HmgA/gCywEvgVKA9WAzcCztvZDgL22fioBawABfLL095TteV8gAv3jrYBGwLXZdbK9rg2cBe5B31h1t72uarseBHwKlARuBeKBn3J5v48CMcBw9KzCO9v1SzraXlcCzqFnIz5Af9vryrbrS4FfgYq2z+c22/kuQLjt+XvA9kx9c9DpcWC97bkP8JLtPZS39ZMGfGx7f/7Z+vYGdgKf2f4mfsAttmv3AQeB6239vgNsdPd3zhwF/J91twLmuDoP2w/zBSAWOAZMBvxt1wKBD7K0rQ4kZ163nesPrLE9/xsYkuVajzyMxQrgpTx0ymos3gBmZWuzAhiEnoWkAaWzXJudm7GwXR8ArAYSbEbnjSzXshuLgcDmbPJBth/4mkAGUDGHMbqgjeGnwHqgfB76PG57D7HAGeBf/jPgXYAUwC9b35nGoiMQnfkZZ+t3OfBkltde6Bnite7+3pnD+mHWIQ3u5H4RWZ3LtRNZnl+LvnuOVEplnvPK0qZWtvbH8hjzGuCQnfpdC/RVSvXKcs4XPXOpBZwTkYRs416TW2ci8jPws1LKF7jf9nyHiKzIoXktrnwfx9CznWuAGBE5l8tQFYBngIdEJC7Xd6f5V0RuyeVatIgk5XLtGuCYiKTlcO1a4Aul1MQs55RN97z+NgYPxvgsDJ5K1nTIJ9AziyoiUsF2lBOR5rbrkVz+I103j35PAA3tGDOz7awsY1YQkdIiMs42ZkWlVGk7x/1vEJFUEfkN2IXNT5PD2CfRP7pZqYueNZwAKimlKuQyxDmgJ9rPcLM9OuWmah7XTgB1c3F8n0AvEWb93PxFZGMBdDG4GWMsDB6PiEQCK4GJSqlySikvpVRDpdRttiZzgReVUnVsO3lG5NHdD8BrSql2tp1WjbI4X08DDbK0/QnopZS60+ZE97M5eeuIyDFgKzBKKVVCKXUL0ItcsG1TvVcpVdam/91on8ymXMZeBlynlHpEKeWjlHoIaAYssX0ey4HJSqmKSilfpdSt2T6zQPSy1wKlVIc8Pg+rbEYbzHFKqdK2zybTMH0DvKmUam577+WVUn1doIOhEDHGwlBUeAwoAexG3znPQ6/dA3yP9iXsRDt0F+TWie2OfizavxAP/I52JgN8BLxj2/n0moicQDtr30Kvz59AO6gz/28eAQLQjuv3gZl56H/e1s9xtI/gE2CoiKy3Xf8C6GPbPfSliJxFzw5eRfs3Xgd6isgZW/uBQCrasR8FvJzDe10FPIHe4dU2D90cRkTS0caxke09hQMP2a4tRDvGf1FKnQdCgbudOb6h8FEipviRwWAwGPLGzCwMBoPBkC/GWBgMBoMhX4yxMBgMBkO+GGNhMBgMhnwplkF5VapUkXr16rlbDYPBYChSbNu27YyIVM3pWrE0FvXq1WPr1q3uVsNgMBiKFEqpXCPszTKUwWAwGPLFGAuDwWAw5IsxFgaDwWDIF2MsDAaDwZAvxlgYDAaDIV+MsTAYDAZDvhhjYTAYDIZ8McbCYDAYDPlijIXBYDAUE/r91o9RgaNc0rcxFgaDwVAMSMtIY/H+xcQl51d23RrGWBgMBkMxYN+ZfSSlJdGmRhuX9G+MhcFgyJ8TJ9ytgSEfgk8FA9C2plMr6F7CZcZCKfWjUipKKRWa5VxfpVSYUipDKdU+W/s3lVIHlVL7lFJ3Zjl/l+3cQaXUCFfpazAYcmHdOqhbFxYudLcmhjzYHrkdPx8/mlRp4pL+XTmzmA7cle1cKNAbWJf1pFKqGfAw0NwmM1kp5a2U8ga+Rhd7bwb0t7U1GAyFxYIF+vGddyA93b26GHKlX/N+TLp7Ej5erkkm7jJjISLrgJhs5/aIyL4cmt8H/CIiySJyBDgIdLAdB0XksIikAL/Y2hoMBgfYHb2bkYEjyZAMxwRFYMkSqFIFdu+GX35xjYKGAnNTnZt4qu1TLuvfU3wWtYGsi6LhtnO5nb8CpdQzSqmtSqmt0dHRLlPUYCiKjFs/jlFrRzE7ZLZjgvv3w6FDMHIktGqlH1NTXaGioQCcTTzL6sOrSUhJcNkYnmIsCoyIfCci7UWkfdWqORZ6MhiuSlLSU/hj3x8AjN84HhGxX3jpUv3YsyeMHg0HD8LMmS7Q0lAQ1hxdQ/dZ3dkdvdtlY3iKsYgArsnyuo7tXG7nDQaDnfx95G/ikuOY0H0CawatQSllv/CSJdCiBVx7rTYYHTrABx9AcrLrFDY4THBkMN7Kmxuq3+CyMTzFWPwBPKyUKqmUqg80BjYDW4DGSqn6SqkSaCf4H27U02AocnSu25l5fecxrMMwKvlXIkMySExNzF8wLg7++QfuvVe/VkrPLo4fh6lTXau0wSGCTwXTrGoz/Hz8XDaGK7fOzgGCgCZKqXCl1JNKqQeUUuFAR2CpUmoFgIiEAXOB3cCfwDARSReRNOB5YAWwB5hra2swGOykdInSPNjsQfx8/EhNT+WmH27i9VWv5y+4ciWkpekZRSbdu0PnzjBmDFy86DqlDQ4RfCqYNjVdE4yXiSt3Q/UXkZoi4isidURkqogstD0vKSLVReTOLO3HikhDEWkiIsuznF8mItfZro11lb4GQ3FkS8QWxq4bS1ySTgHh6+3LjbVu5Jut37DvTE4bE7OwdClUrAg33fTfuczZRWQkfPONCzU32MupC6c4deGUyyK3M/GUZSiDweACpu+Yzth/xl629/79Lu9TyrcUb6x+I3fBjAxYtgzuvht8su3bv+026NYNPvoILlxwkeYGe6lSqgrbntlGv+b9XDqOMRYGQzElQzJYuHchdzW6i9IlSl86X610Nd685U0W7VvE2qNrcxbesgWio//zV2Rn9Gh9fdIkF2hucAQfLx/a1mxLrbK1XDqOMRYGQzEl6EQQkRciefD6B6+49vJNL1OnXB0mb52cs/DSpeDlBXdlT8Jg46abtCEZP147wg1uY/qO6Szet9jl4xhjYTAUU+bvmY+vly89r+t5xTV/X39WDVzFrAdm5Sy8ZAl06gSVKuU+wAcfwLlz8NlnTtLYYIXR60Yza1cuf0cnYoyFwVBMSUhJoOd1PSnvVz7H602rNKWEdwkSUhJISkv670JEBAQHX74LKifatoUHH4RPP4WzZ52oucFeYpNiOXzusMud22CMhcFQbPm217fM7zc/zzbRCdE0ntSYLzd9+d/JZcv0Y27+iqyMGqWd3OPHF0BTg1V2nNoB4PJts2CMhcFQLLmYqmMg8ovWrlq6Ku1qtWPsP2M5k3hGn1y6VEdsN2+e/0DNm0P//trRffp0QdU2OEhwpK5hYWYWBoPBYUSEtt+15eU/X7ar/cfdPuZCygU+WPsBJCXBqlV6VmFvWpD339fpP8aNK4DWBivsP7ufWmVrUb1MdZePZYyFwVDMCIsOY++ZvTSt0tSu9s2qNuPptk8zZesUwhf/DImJ9i1BZXLddTBoEEyZAuHhFrU2WGHyvZMJe65wkloYY2EwFDPm756PQnF/0/vtlhnZZSR+Pn5EzP4W/P3h9tsdG/Tdd3Ug31iTZKEwUUpRwa9CoYxljIXBUMyYv2c+N9e9mRplatgtU6NMDXY8E0yHnWfgjju0wXCEevXgqad0gsEjRxyTNVhi1+ldPDL/EQ7GHCyU8YyxMBiKEQfOHiAkKiTHQLz8aBiVijpyhLhunR2vqAfw9ts6kG/0aMdlDQ6z8cRG5oTOcVkZ1ewYY2EwFCOqlKrCV3d/Rd9mfR0XXrIEgPbh7zE3bK7j8rVrw9ChMGOGrrBncCnBkcFU9KvIteWvLZTxjLEwGIoRFf0rMqzDMGqXy7H6cN4sXYq0akXphk0ZsXrE5YF69jJiBPj56fgLg0sJPhVM6xqtHStmVQCMsTAYigkR5yOYun3qpXTkDnHuHKxfj7r3Xib0mMCxuGN8tfkrx/upXh1efBHmzIHQUMflDXaRlpFGSFRIocRXZGKMhcFQTJgbNpenFj9FdGK048IrV0J6OvTsSbcG3bi70d2MWTeGs4kW0ni89hqUKaPjLwwuISohiiaVm9ChdodCG9MYC4OhmDB/z3xaVm9Jo0qNHBdesgSqVNE1toFPun9CYmoiKw+tdLyvypXhlVdgwQLYvt1xeUO+1Cpbix1DdvBQi4cKbUxjLAyGYkBkfCQbT2y0tAuK9HRYvlynI/f2BqBFtRYcffko/W/ob02h//1PV9l77z1r8gaPwxgLg6EYsHDvQgSxZiw2b9ZZY7Nlmc0spnMo5pDjfZYvD8OH6zxTQUGOyxvy5L5f7uP5Zc8X6pjGWBgMxYCQ0yE0rdKUZlWbOS68ZImeUdx55xWXFuxZQONJjdl4YqPj/b7wAlStqqO7DU4jQzIIPBpoLRamABhjYTAUA6b0nMLmpzZb20a5dCnccgtUuDJtxJ0N76RGmRq8uvJVRMSxfsuUgTffhL/+gtWrHdfLkCNHzh3hfPL5Qt0JBcZYGAxFnswf8bIlyzoufOIE7NyZa+LA0iVKM6brGP4N/5efQ352vP8hQ6BhQxgwAI4dc1zecAXBp2xpyQuhhkVWjLEwGIo4/eb148XlL1oTzix0lEdVvMdbP05A7QBeWfEKMRdjHOvf3x8WL9YpzHv1gvh4a3oaLrE9cjs+Xj60qNaiUMc1xsJgKMLEJcWxaO8iSniXsNbBkiVQvz40zT2duZfy4rte35GSnsK2k9scH+P66+G332D3bl0oKT3dmq4GAJpUbsLTbZ/Gz8evUMc1xsJgKMIs2b+E1IxUa7ugLl7U/oSePfMtdNSyektO/O8E3Rt2t6Zo9+66mt7SpXqXlMEyg1oPYvK9kwt9XGMsDIYizPw986lVthYBdQIcFw4M1AbDzkJHZUuWRURYsGcBKekpjo83dKhOBfLZZ/Dtt47LG0hOS+ZCygW3jG2MhcFQRElISeDPg3/Su2lvvJSFf+UlS6B0abjtNrtFgsKDeHDug4zfMN7x8QAmToS774Zhw8wOKQv8deQvyn1Ujs0Rmwt9bGMsDIYiSkp6Cq92fJWBrQY6Liyil4S6ddNZYu2k0zWd6NusL6PXjbZWdMfHB375RftI+vSBvXsd78NOziSeYcepHS7r3x1sj9yOIHaXzHUmxlgYDEWUiv4VGd11tLVkcmFheiurI7W2bXx+1+eU9CnJ0KVDHY+9AChXTs9qSpTQ/pKzFpIV2sH55PP0+60fI1aPIDU91TFhEYi2kJDRxQSfCqZRpUaUK1mu0Mc2xsJgKIIkpSWxdP9SktOSrXWwdKl+vOceh0Vrla3FR3d8xOrDq5kdMtva+PXqwe+/Q3g49O4NKRZ8ILkgIogINcvUpGv9rny84WM6T+vMkXMOlHt97TWoVk3nuLp40Wm6FZTtkdtpW7OtW8Y2xsLgPtasgX//dbcWRZKVh1bSc05P1hxdY62DJUugTRtd3c4Cz7Z7lt7X96aC35VR33bTqRP8+COsW6eD96zMUnLgq81f0ee3PgB80/Mb5vaZy54ze2j9bWt+C/st/w6++AI+/RTatYPPP4f27SE42Cm6FYRzF89xNPZooUduZ2KMhcE9rFsHPXrodeu0NHdrU+SYv2c+5UuWp2v9ro4Lx8TAxo15BuLlh7eXN/P7zefe6xxfxrqMRx7RmWmnTYPxFp3mWdhwfAOvrHyF1PRUSvqUBKBv877seHYH11e5nuUHl+fdwcKFejbxwAOwaROsWAGxsTp1+9ixbv2ueikvvrzrS+5tXMDP3CqZU7bidLRr104MHszx4yJVq4qULy8CIgsXulujIkVyWrJUGFdBBi4YaK2D2bP15/7vv07RZey6sfLPsX+sd5KRIfLQQyJKiSxYYLmbyPhIqTmhpjT8oqGcu3juiuspaSmSmJIoIiKhp0Ml9HTo5Q2CgkT8/EQCAkQSEv47f/as1g9EOnYUOXDAso6eDrBVcvldNTMLQ+Fy8SLcfz8kJcH69VCnDkyZ4m6tihSBRwOJTYq1FogHegmqalW48cYC65Kansp3277j2SXPWou9AB0QOG2a1ufRRy0VTEpNT+WheQ8RmxTLgocW5Lg85uvti7+vPwDPL3+eG7+/kR+2/6Cd9IcO6XQktWvr9CSlSv0nWKmS3sE1ezbs2QOtW8N33zlt2cxetkRs4Vis+/JrGWNhKDxE4Omn9frvzz9Dixb69cqVcNDCNsyrlBUHV1DatzQ9GvZwXDgtDf78Uzu2vQr+71+6RGkm3zuZ3dG7mbBxgvWO/P1h0SJdZa9XL4iIcEj80LlD7Inew/e9vqdl9Zb5tp/dezadrunE04uf5plpvUm/6079/Vy2TBvSnOjfH0JC4Kab4NlntZ6nTjmkZ0F4fNHjPL+8cGtYXEZuU46ifJhlqMIhPjleXl/5umyJ2GKfwMSJeio/evR/5yIiRLy9RV57zTVKFkPSM9Jl/5n91oT/+Uf/DX77zak69ZnbR/zG+MmBswVcotm5U6RMGZG2bUUuXHBINKelp7xIz0iXT1aNkg3XIEk+SsKXz7VTMF3kiy/0klXlyiLz5zs0rhUSUhLEa5SXvPv3uy4dhzyWodz+w+6KwxgL13Ms9pi0mtJKGIm0/659/gIrV4p4eYn07q3/2bLy4IP6n+7iRdcoa/iPESNEfHxEYmOd2m3E+Qgp+2FZuXPWnQXvbPFi7b/I6buSjbCoMBmzdoykZ+TdLkfS00X69JEMpWTCKzdJclqyY/K7d4u0a6d/RgcNcvpnmpV/T/wrjEQW7nGtf88txgL4EYgCQrOcqwSsAg7YHivazpcHFgM7gTBgcBaZQbb2B4BB9oxtjIVr2Xh8o1QbX03Kf1ReRgWOkpPnT+YtcPCgSMWKIi1aiMTHX3l99Wr9VZw50zUKFyNGBY6SV1e8ar2DFi1Ebr/deQpl4bew3+yfZebHp5/q78Sbb+baJC4pTppMaiLVxleTU/GnHB/j1Vf1GBMnXjp1NvGsDP59sJy+cNq+PlJSRN59V98I1a0rEhjouB52MGXLFGEkcvTcUZf0n4m7jMWtQNtsxuITYITt+QjgY9vzt7I8rwrEACVsxuWw7bGi7XnF/MY2xsJ1ZGRkSIfvO0jDLxrKnug9+QvEx+sfqIoVtdHIuVOR664T6dTJucoWM9Iz0qX2xNpy/y/3W+vg2LErfhxdhaU7/axkZIg884zW9+OPRZKTs13OkN6/9hbvUd6y5sgax/ufNEn3/fzzeiwby/Yvk5KjS0rNCTXlr8N/2d9fUJBIo0Z6RvTqq06fJT/9x9NScVxFyciiqytw2zIUUC+bsdgH1LQ9rwnssz1/E5gMKKA+cBDtfO8PfJtF/lugf37jGmPhfNIz0i9tOzwRd0LOJJy5dO1wzGG5ddqtsvH4xsuFMjL0EpOXl16GyovMO8mdO52tukfRZ24fafhFQ7nph5uk5+yel2ZlWyO2yo/bf5TF+xZL0IkgOXD2gJy7eO6yH4egE0HCSGTWzlnWBp88WX/Ge/c6463kSEZGhgxdMlSe+eOZgneWkiLSs6fWuX59kenTRVJTRURk/IbxwkhkwoYJjvf7++/6O/l//yeSlnbF5R2RO6TJpCbiNcpLFux2YCvvhQsiQ4ZofVu0ENljx82UnRyLPSaBR1wza8mKJxmL2CzPVeZroCywBogELgD32s6/BryTReZd4LVcxnoG2ApsrVu3rqs+y6uS+OR4uW/OffLALw/keMcYnxwvlT+uLL1m97r8wujR+is2wY5/6LNntcNwyBAnae15nEk4I4xE2n7bVrrN7Catv2l9yei+tfotYSRXHOeTzouIyGdBn0nDLxqK7we+DjtyL3HPPfru18V3p6+teE0YScFiLzLJyBBZulSkTRv9XWrSRM5NmyL+H5SUvnP7On6nvWmTiL+/SIcOl8dSZCM+OV4Cvg8QvzF+suH4BsfGWLr0vziiFSsck3UzHmksbK/P2R77AJ/ZDEgj4AhQzhFjkfUwMwvncfTcUWk5paV4jfKSSZsm5frPOSpwlDAS2XVqlz7xxx96Sj5ggP0/ToMG6Z0w5887R3kPY+XBlcJIclw2SUhJkCPnjsjm8M2ybP8ymbFjhkzcOPHS5/3lv1/KLT/eIm+tfsva4AkJ2hi/9FIB3oF9XEi+IHU/qyvNvm7muNM4NzIydMBe8+YiIAlNG0ri3NmOGb5Dh/SPeP36Iqfz90lEXYiSvnP75u+Ty4mjR0VattQzmC+/LJCBPnruqEzaNMl+P0oB8CRjkdsy1FKgc5Z2fwMdzDKUe9lwfMMlR/aKg3nfIZ1NPCtlPiwjj8x/RE+/y5bV2x8TE+0f8N9/9Vdy8uQCau65hMeFS1JqUuEOeviwyHvv6c921apCGXLxvsXCSGTM2jFO6zM5LVlW7f9TR6A3bqzfT/v2IsuX5/9jfOaMSJMmIpUqWVqGS01PlagLUY4JxcfrpS4QefZZvaxmgenB04WRyO6o3ZbkHcGTjMX4bA7uT2zPpwAjbc+rAxFAFZtj+4jNuV3R9rxSfuMaY1FwklKTpM6ndaTRl43sc2SLyPCVw6XiCCXJDevrO7hjxxwbNCNDLzfccIPLl0qKNadPi8yZI/LUU/ouWoeb6c82qfAMVZ+5faTKJ1XkQrJj8RK5MWzpMGEkEnI6RPsufvxRpF49/d46dRL5+++cBS9eFLnlFpGSJXWciQX6z+svN0y+wfElwPR0kTfe0Dp27aqXWx3kxWUvSqmxpSQt/Ur/irNxi7EA5th8EKlAOPAkUBn4C70NdnXmDz9QC1gJhAChwKNZ+nnC5vA+SJYttXkdxlhYJz0j/ZJfYtvJbXI20f4v98lzJ+RQx+slw8dHZO1aawp8953+Wq5fb03eQ0lLT5MB8wfI34dz+UErCHFxOjbh5Ze1oc00DuXLi9x3n14GCQsrdAMcGR8px2IdvGHIhZk7ZgojkddWZAveTE4WmTJFpFat/36QN2TxMaSni/Trp6/9+qvl8VceXCm+H/jKbdNuk4upFnY6zZghUqKE9hk56Pju/GNn6fhDR8fHtIDbZhbuOoyxsMb5pPPSa3YveX/N+9Y6eOst/ZX6+mvrSsTHi5Qrp30dxYgdkTuEkchPO38qeGdJSfou+p13dGI7b2/9ufv5iXTrJvLRRyKbN1/aOeRuElIS5LZpt8njvz8uHwR+ID/t/Ek2HN8gMYkxdsnviNwh/mP8pcv0LpKanst7SkwU+ewzkWrV9Gdxzz0iW7eKDB+uX3/ySYHfx8+7fhZGIn3n9rW2NXjDBocd3+kZ6VL2w7Ly3JLnHB/PAsZYGPLlyLkjcsPkG8R7lLd8tekrxzuYO1d/nZ56SmYGz5BPN35qXZlhw/RdWHS09T48jMygqkMxh6x3sn69NgZ+fvqz9vYWuekmkbff1sbDQyPgT54/KTdPvVlqTax12U6vz4I+ExH93bvn53vk+aXPy8SNE2XhnoWyI3KHJKQkSEJKgjT8oqHUmljLvsC7+HhtLCtWlEszrOeec9qsasKGCcJIZPjK4dY6OHpUz/68vXWsRz56HT13VNRIJd9v+97aeA5ijIUhT9YfWy9VP6kqFcZVkJUH84mHyImdO0VKldJ3uUlJ8uiCR6X02NKXxWI4RGio0+4GPYXHFj4m1cZXK1hQVfv2+s75pZf0brO4OOcpWEgkpiTKnug9smz/skuGMzgyWFpNaSXlPip3mTGZFzZPRER+2PaD49tXY2NFRo3Sn5WTZ1gj14wsWKT6+fP/Ob6HDMnX8X0+6bzEJ+eQ+cAFFMhYAKVsW1a/t71uDPTMT86dhzEW9nPu4jkp82EZafxlY9kbbSFY68wZ7UStVUvkpN5iGHo6VBiJ9eUsEZHOnUUaNsw3N1BRofGXjeW+OfdZ7yA2Vm/DfO895ynlYWRkZMjZxLOyJWKLzA2da23LaiGz78w+a4JpaSKvvy4FcXy7gryMhT05iqcByUBH2+sIYIwdcgYP5mziWQAq+FVg1gOz+Pepf2lSpYljnezdq2tTRETAggVQsyYAzas1574m9/Hlpi+JT463puDQobrGwKpV1uQ9iKS0JMqWLEvnup2td/LPP5CRAV26OE0vT0MpRSX/SrSv1Z6+zftSs2xNd6uUJ9OCp9F8cnNWHFzhuLC3N3z8MUyfruu6BATAvn1XNHtvzXtM2eIh9V5ysyKZBzZLAwRnObczPzl3HmZmkTsZGRkyLdec4QYAACAASURBVHialP2wrGOpDLKye7dI//466K5UKZGfrnTaZmbJtJSOQUQ7catW1bt5DDrfUMmSHuuXuBqJvRgrraa0ktJjSxdsWWr9+v8c31nS4mRkZEi18dXk8d8fd4K29kEBZxYpSil/QOfoUKoheqZhKGJEJ0TTe25vBi8aTJuabWhT08HC77t36wIwzZvrQjWvvQZHjsCAAVc0DagTwHPtn6Nx5cbWlC1ZEp58UlctO3HCWh/FicBAXXTHz8/dmhhslPcrz/IBy6lSqgr3zr6XQzGHrHV0882weTPUrQt33w1ffw1A5IVIohKiaFujrRO1LgC5WZHMA+gOrAWigZ+Bo0CX/OTceZiZxZUs279Mqo2vJiVGl5DxG8Y7FuATEqL3qislUrq0DjKKcjCa1QqHD+sx33VtwRdX02t2r4JtfTx3Tvsr3n/faToZnMfe6L1S+ePK0vjLxtZiMDI5f16kVy/tx3jsMVm+9Rfn5diyEwoysxCRVUBv4HF0oF17EQl0ieUyuIwziWeoUaYGW5/eymudXsPbyzt/oZAQ6NcPbrhBl5scMQKOHoVx43IvPZmN2KRYvvj3C9Iy0hxXun59uOsu+OEHSE11XN4DSMtI468jf+GlClDC9CrwVxRlmlRpwpJHlvD+be/j51OAmV/ZsrBwIbz3Hvz8Mx3veZZuh6BV9VbOU7Yg5GZFMg/gAaB8ltcVgPvzk3PnYWYWmvXH1svPu34WEb3+mZJmZ26anTt1anHQOZ7eekvverLA73t+L1gw2h9/aD3mzbMm72a2n9wujOTS38ESr7xi/BVFiG0nt9n/v5YbmzZJVN3K+rs/bJjDJWatQgG3zu7I4VxwfnLuPK52Y5GcliwjVo0Qr1Fe0vzr5rlHvWZnxw5dyjLTSLzzToG39KVnpEvzr5tLi8ktrEW9pqXpCmR33FEgPdzF15u/FkYih2MOW++kbVuRLl2cp5TBZRw5d0RKji4pgxYOKnihosREkf/9Ty/FNmxYKClw8jIW9syNc2rj44xZjcH5hEaF0uH7DozbMI7BrQcT9GQQPl75/Ll27IAHHoDWrWH1anj3Xb3cNHo0VKpUIH28lBcjbhlBaFQoS/YvcbwDb2945hn46y/Yv79AuriDoPAgapSpQb0K9ax1cO4cBAebJagiQr0K9XjzljeZsXMGd/50J0fOHbHemb8/fPoprFkD6enQuTO8/jokJTlPYUfIzYpkHuha2p8CDW3Hp8D0/OTceVytM4vwuHDxG+MnVT+pKov2LrJPaN8+ER8fvW3v/fdFYuzL1+MIqempUu/zehLwfYC1u63ISK3j//7ndN1czVebvpJRgaOsd7BokZ7pWU3MaCh0MjIy5OvNX0vZD8uK/xh/+eLfLxzuY+3RtdLmmzYSFhWmT5w//1+Z2WbNdN4rF0ABZxYvACnAr7YjGRjmCsNlKBi1y9Xm63u+JvS5UP6vyf/ZJ7RhA6Sl6cCgkSOhYkWn6+Xj5cMbN79BBb8KxKdYCNKrUQN699YBTBcvOl0/VzKswzDeu+096x0EBurtsh06OE0ng2tRSvHcjc8R9lwY3Rp049zFcw73sSViC8GngqlayraRpGxZ+PZbWL4cYmP1NupRowp340duVqQoH1fbzOKPvX/Ix+s/tnbX/r//6cR0OdQidiYFXr/9+299VzVtmlP0KQyiE6IlLqmA+ZvatBG5/XbnKGQodDIyMi5tU1+8b7G8seqNS7Xs8+KR+Y9InU/r5HwxJkZnZQbtzwoNdZq+WJlZKKU+tz0uVkr9kf0oNGtmyJc5oXOYvGUySinHhUNCoFkz7RtwIZm6HYs9xsGYg4530KULNG0KUzwk9YEdTNg4gWrjq5GcZjGGNSZG+5OMv6LIopS6tE096EQQH2/4mBum3MDfR/7OUy44Mpg2NXIJmq1YEX76CebNg+PHoV07mDBB+zVcSF7LULNsjxOAiTkcBg8hNCqUFtVaWBMOCdFxFIVAanoqHX7owPBVwx0XVgqGDNGRrtu3O185FxAUHkTL6i0p6VPSWgf//KOTbN9+u3MVM7iFsXeM5a/H/gLgjpl38MSiJ4i5GHNFu8TURPad3Ze7scjkwQchLExHfQ8fDrfdBgct3IjZSa7GQkS2KaW8gWdEZG32w2UaGRwiNT2VvWf2WjMW0dFw+nShGQtfb1+GtBvC73t/Z3f0bsc7GDRI7xApArOL1PRUtkRsoWOdjvk3zg3jryh2dK3flZChIYy4eQQzd85k6f6lV7Q5n3yefs37ceu1t+bfYbVqOonnzJkQGgqtWsHkyfomw8nk6eAWkXTgWqVUCaePbHAKB2MOkpqRas1YhITox0IyFgAvBrxIKd9SjFs/znHhChV0bqrZsyEuzvnKOZFdp3dxMe0iHa8pgLFYswY6ddJ5sgzFBn9ffz7q9hFhz4XxaMtHAVh2YBkn4nQOtBplajDnwTnc0eAO+zpUCgYO1MbillvgD9d4CezZDXUY2KCUelcp9Urm4RJtDA5zNPYoPl4+NK/a3HFhNxiLyqUq82y7Z5kdMtvaHvShQyExEWbNyr+tGwkKDwKwPrOIiYFdu8wSVDGmSZUmKKVITkvmyT+epNnkZny1+Svraf3r1IE//4TfftMGxMnYYywOAUtsbctmOQwewN2N7ybhrQRaVm/puHBICFSurLemFiKvdnwVPx8/1h9f77hw+/b6mDLFJVNtZ3FXo7v4tue31C1f11oH69bp92ec28Wekj4lCXoyiJuvuZkXlr9AuXHl+L85dm59z45SeputC8gztFcpVRVYChwUkViXaGAoMCW8La4SZjq3XXAXkhe1y9Um4pUIyvuVt9bBkCHw1FPaAXyrHeu6bqBRpUY0qtTIegeBgdo/c+ONTtPJ4LnUq1CP5QOW83PIz7y4/EVuu/Y2d6t0BXltnX0KCAMmAXuVUhZNncGVDPp9ED8G/+i4YEaG3klRiEtQWck0FDtO7SA13cHAoocfhvLlYdIkj5xdxFyM4ZfQX3Lc6WI3xl9x1aGU4tGWjxLzRgyvdnrV3epcQV7LUC8DzUWkI9AJeLNwVDLYS1JaEj/t+oljscccFz5yBBIS3GYsAGbunEmbb9vQcWpHwqLC7BcsXRqefVbvM+/SxeO20q47to7+8/uz98xeax2cPWv8FQaPIy9jkSIi0QAichgwtzgext4ze8mQDJpXKxrO7ew81uoxfuv7G8fijtH2u7Z8vP5j++tejB0L33yjq/e1bw9PPAGRka5V2E6CTgTh6+VL25oWK5ytW6cfjb/C4EHkZSzqKKW+zDxyeG1wM6FRoQAF2zbb3IKhcSJ9mvUh7Lkwel7XkxF/jeD+X+63T9DHR88uDhyAV1/VEa2NG8OHH7o9f9TG8I20rdnWeiGcNWuMv8LgceRlLIYD27Ic2V8b3ExoVCi+Xr40rmShznVIiK5E56KdE45QrXQ15vWdx5wH5/BkmycByJAM0jPsSF9QoQKMH69nGN27w9tvw/XXw6+/usWfkZKewtaTWwsejHfzzVDChDcZPIdcd0OJyIzCVMTgOCW8S9ClXhd8vX0dFy7ENB/2oJTi4RYPX3r9+b+fs3DvQqbdN82+XUWNGumSlGvWwP/+p53gkybBZ58V6h16yOkQktKSrAfjnTmj/zYPP5x/W4OhEClAYWCDu/ng9g9YOXCl44JJSXr5xoOMRXZqlKlBaFQoLae0ZNKmSWRIhn2Ct98O27bB99/r99ihg04TEhHhWoVttK3ZliMvHeHuRndb68D4KwweijEWVyN79ugMlR5sLB654RFCh4bSpV4XXvzzRbrO6Gp/xLe3t47DOHAA3ngDfvkFrrtOV/5LTHSp3kop6lWoR9mSFpf31qyBUqW0095g8CDyNRZKqSvqaiql6rtGHYO9bInYQpOvmrA5YrPjwh6wE8oeaperzdJHljL1/6ay8/ROws+HO9ZBuXIwbpw2jvfcA++9p9Ocz5njMn/Gi8tfZMXBFdY7CAzU+X2Mv8LgYdgzs1islCqX+UIp1QxY7DqVDPYQEhXC/rP7qexf2YJwiP4xamzBMV7IKKV4os0THHv5GJ2v7QzA1O1TOR533P5OGjTQ+XLWroUqVeCRR3Tk94ULTtU1Mj6SSZsnWcuoCzoLcGioWYIyeCT2GIsP0QajjFKqHfAb8Khr1TLkR2hUKP4+/tSvaGGSFxKidwz5WnCMu4lyJfX9ypnEM7yy8hVaTG7Bz7t+dqyTW2+FLVvgq690GdlffnGqjpeSB1p1bq+1Zf43xsLggeRrLERkKfAZsBKYDjwgIjtcrJchH0KjQmlWtRleyoLbycN2QjlClVJV2PHsDlpWb8mg3wdxMv6kYx14e8Nzz2ljOW2aU3ULOhFECe8S+RetyY3AQB2dbvwVBg8kr9xQk7IE4HUFygNHgOdNUJ77sVwdLyYGTp4sssYCoH7F+ky/fzrpks7U7VMd70ApGDwYNm6EffucpldQeBDtarazXhkv019RhGZ8hquHvG5Lt3J5EN4nwHxMUJ7bSctI465Gd9GjYQ/HhYuIczs/GlVqRI+GPfgh+Af7t9VmZeBAPcuYPt0p+ogIqRmp3HzNzdY6iIrSiR3NEpTBQ8k3KE8pVRpIslXNw1Zq1eSJciM+Xj78eJ+FTLNQbIwFwMQeEyntW9raUlyNGrp28cyZMGaMNhwFQCnFpqc2IVZ3WRl/hcHDsee/7C/AP8trf2B1fkJKqR+VUlFKqdAs5yoppVYppQ7YHitmudZFKbVDKRWmlFqb5fxdSql9SqmDSqkR9r2t4k1iaqK1u2nQxqJCBahd27lKuYEW1VpYc/BnMniwXpJbaSGwMReU1doggYFQpgy0a+c0XQwGZ2KPsfATkUt7DG3PS9khNx24K9u5EcBfItIYbYRGACilKgCTgf8TkeZAX9t5b+Br4G6gGdDftnX3qmb4yuFc+/m11u5i3VTwyFXsO7OPXnN6cSjmkOPCPXvqrbROcHQ/u/hZBi8abL0D468weDj2GIsEpdSlXMu27bP5pvUUkXVA9uov9wGZOadmAJkpRh8BFojIcZtslO18B3SVvsMikgL8YuvjqiYsOoy65es6fhcrovfxF4MlqEzKlCjD8gPL+Xbbt44LlygBAwbAokW6hkQBWH5wORdTLWa7PX1aJ0I0S1AGD8YeY/Ey8JtS6h+l1HrgV+B5i+NVF5HMogOngOq259cBFZVSgUqpbUqpx2znawMnssiH285dtYiI3glV1cJOqGPHID6+WBmL2uVqc1/T+/gx+EeS05Id72DwYEhJgdmzLesQcT6CE+dPWM80a/wVhiKAPXEWW4CmwFBgCHC9iBR4N5ToNZTMdRQfoB1wL3An8K5S6jpH+lNKPaOU2qqU2hodHV1Q9TyW0wmnOXvxbJEteOQKhrYfytmLZ5m3e57jwq1aQZs2BVqKKnAwnvFXGIoA9uSG8kUbipG241nbOSucVkrVtPVbE8hcbgoHVohIgoicAdYBrYAI4Jos8nVs565ARL4TkfYi0r5q1aoW1fN8nFLwqIUFWQ+ma/2uNKrUiClbp1jrYPBgCA6GnTstiQedCMLPx4/WNVpbGz8wEDp31gWdDAYPxZ5lqCnou/7JtqOd7ZwV/gAG2Z4PAhbZni8CblFK+SilSgEBwB5gC9BYKVVfKVUCeNjWh2uIjoaXXoING1w2REG5ptw1vHvru7Sq3spx4ZAQqFsXypd3vmJuxEt58Xbnt7m70d3Wdok98oj2X1icXTSt0pSn2z5NCW8Lyf9OndKJDs0SlMHTEZE8D2CnPedyaDMHiARS0TOHJ4HK6F1QB9DbbytlaT8c2A2EAi9nOX8PsB84BLyd37giQrt27cQS8fEiXl4i775rTd7Tad5c5N573a2FZ9K3r0iVKiLJyYU77i+/iIDIpk2FO67BkAPAVsnld9WemUW6Uqph5gulVAMg33qXItJfRGqKiK+I1BGRqSJyVkTuEJHGItJNRGKytB8vIs1EpIWIfJ7l/DIRuU5EGorIWDv0tU6ZMnqJZtMmlw5TEHae2klcUpzjgikpOrVFMfNXZCUlPYW5YXOJT453XHjwYF2lbskSh8TOJ58nMbUANTICA3Vp27Zt821qMLgTe4zFcGCNbafSWuBv4DXXquVGAgJg82bIsBj05kJEhFum3cI7f7/juPDevZCWVqyNxfbI7Tw07yFmh1jY2dSjB9Sq5fBS1A/bf6DcR+U4k3jG8TFBFzsy/gpDEcAeY7EeaAy8CLwANAE8d1G/oAQEQGysrrLmYRyPO86FlAsFc24XY2MRUDuAVtVbMWXrFMcDFr294bHHYPly7Uewk6DwIOqUq0OVUlUc1BaIjNSzPeOvMBQB7DEWQSKSLCK7bEcyEORqxdxGQIB+9MClqALvhPLxgSZNnKyV56CUYkj7Iew8vZNNERb+foMH63Kzs2bZLRJ0IohO13RyfCz4L77i9tutyRsMhUheKcpr2KK1/ZVSbZRSbW1HF+xL91E0uf567bvwQGMRFh0GYD3GomnTYl+uc8ANAyhTooy1bbTXXQedOumlKDtmJifiThARH2E9GC8wUJd+bW1xy63BUIjktVB6J/A4OrZhIpCZWyIeeMu1arkRb29dfMYDjUVoVCi1y9amgl8Fx4VDQnTuoWJO2ZJlGdhyIEHhQaRnpOPt5WA22cGD4emn9d//ppvybFrgYDzjrzAUIXKdWYjIDBG5HXhcRLqKyO224/9EZEEh6lj4BAToAK2LFnP9uIhXOr7CNz2/cVwwNhZOnCjW/oqsjO8+nu3PbHfcUAD06wf+/nY5utvXas/EHhOtxbycPAn79xt/haHIkNcyVC+l1LUiMt/2+j2l1E6l1B9KqQLkhS4CBATonUM7PKt6bOsarel5XU/HBUNtWeKvEmNRukRplFLWUrmXKwd9+uj63Il5b4ltULEBr3R8BV9vCwkNjL/CUMTIy8E9FogGUEr1BB4FnkBHUFu4vS1CeKCTOyohirlhc4m5mD2Rrx1cBTuhsrP15FZqTazF30f+dlx48GA4fx4WLsy1SVJaEvN3z+dsosVstWvWGH+FoUiRl7EQEcm8teoNTBWRbSLyA1B8ky+B3m9fp45HGYsNxzfw0LyHOHzusOPCISH6h6luXecr5qG0qNYCHy8fa47u226DevXyXIraHrmdPr/14Z/j/1hTMDAQbr21wBX6DIbCIi9joZRSZZRSXsAd6DQdmfi5Vi0PICDAo4xF5rbZ66tc77hwSIiOTC8mBY/swc/HjyfaPMGivYuIOJ9j7snc8fKCxx+Hv//Wad1zYOOJjQDWdkJFROg4HrMEZShC5GUsPgd2AFuBPSKyFUAp1Qad86l4ExAAR47o5IIeQGh0KA0qNqB0idKOCYr8Vx3vKuOZds+QLulMDZ7quPCgQfqzmzEjx8tB4UHUr1Cf6mWq53g9T0z9CkMRJK/dUD8Ct6ETAN6T5dIpoAD1I4sIHua3CI0KtRaMFx4OcXFXpbFoVKkRPRr24Ltt35GWkeaYcL160LUrTJ9+ReoXESlYMN6aNTrzbysLu6gMBjeRZwS3iESISLDIf1tKRCRSbOVPizXt2un1ZA8wFinpKew/u5/mVU3BI0f5sOuH/P7w7/h4WYhleOIJPbtct+6y08fjjhN5IbJgwXjGX2EoYphooNwoXdpjMtD6evly4IUD+HpZ2KJ5lRuLdrUKUH3ugQf0xoBp0y5bMqpbvi77n99PRf+KjvcZHg4HD8Jzz1nXy2BwA/bkhrp68ZAMtEop6lWoR+1yFsqPh4RA7dpQ0cIPWzHheNxxnlz0JIdiDjkmWKoUPPww/Pab3kprQylF48qNrSUPDAzUj8ZfYShi5BWUVymvozCVdBsBAXq9f/9+t6rx+97f+WrzV9aEr1Lndla8lTczds7g223fOi48eLCO5J8799KpUYGjWLLfsboXlwgMhAoVoGVLa/IGg5vIa2axDb0TalsOx1bXq+YBdOigH928FDVj5wwmb5nsuGBqqi7ZeZUbi9rlanNf0/v4MfhHktOSHRMOCNAJGG0xF/+G/8uYf8aw4biFLP3Ll+tAP+OvMBRB8toNVV9EGtgesx8NClNJt+EhGWhDo0KtZZrdv18bjKvcWAAMbT+UsxfPMm/3PMcEldKzi40beeKTm+k4tSPlS5bnkRsesb+PiAjo2xfuuQeqVoXRox3TwWDwAOzyWSilKiqlOiilbs08XK2YR+DtDTfeqP0WbiIxNZFDMYdoUdUUPCoIXet3pVGlRpYiumP79CLNC1r/uYNPun3C0ZePckN1Oz7TtDT44gs9M1myBMaM0QkqzRKUoQiSr7FQSj0FrANWAKNsjyNdq5YH4eYMtHui9yCI9YJH3t56hnSV46W8eK3ja7Sr2S7fmAsRYfmB5QxbOgwRoUKD64nr0pHn95Vn+E2vUKZEmfwH3LJFf3defhluvlknc3z7bShZ0knvyGAoXOyZWbwE3Agcs6UsbwPEulQrTyIzA21wsFuGPxp7FC/lZb3g0XXXmR8oG8+2f5Yv7v4i15gLEWHR3kV0+KED98y+h8X7F3Pqgi6xWvm5V/E6GQkrV+Y9SFwcPP+8/t5ERsKvv2pfRcOGzn47BkOhYo+xSBKRJAClVEkR2Yuuw3114OZI7gebPciFNy9wXeXrHBc2O6GuQEQIPBpIfHL8ZecPnD1A629bc/+v9xNzMYYfev3AwRcPUrNsTd2gVy+oXDn35IIiOq1506YwZYo2GHv26PoYV1FOLkPxxR5jEa6UqgD8DqxSSi0Ccs6uVhypWROuucatTm5/X3+8lIMhMfHxcPSoMRbZ2B65ndtn3M7skNmkZaRdyuJbp1wdKvpVZOb9M9n3/D6ebPskJbyzlKAtUQIGDIBFiyAmW5r4gwfhzjuhf38d07JpE3z5pU7pYTAUE/L9BRKRB0QkVkRGAu8CU4H7Xa2YR+HGDLQDFw5kTsgcxwWvsoJH9tK2Zlta12jNuA3juP7r6+k+qztpGWn4+/oT+HggA1sNzD01yODBkJICs2fr18nJ8MEHOtL/339h0iT9PWnfvvDekMFQSNi7G+oWpdRgEVkLBAEWQomLMAEB+i49KqpQhz2ffJ6fdv3EsTgLEzmzEypHlFI8f+PzHI09StkSZZnQfYL9s7bWrfUxbZpOX96yJbz/Ptx/P+zdq5eeTPyEoZiSb24opdT7QHu0n2Ia4Av8BNzsWtU8iKx+i169Cm3YsKgwAOs7oUqX1tlTDZfxRJsn6HhNR66vcj3KUX/C4MHw0ktwxx3QoAH8+adegjIYijn23FI9APwfkAAgIieBsq5UyuNwUwbazIJHlrPNtmihC/kYLkMpRbOqzRw3FACPPqoj+995Ry/1GUNhuEqwJ+tsioiIUkoAlFIOVt8pBpQqpZdzXGksfv4ZwsLgww8vnQqNCqW0b2murXCtY31lFjzq3dvJShqoVMntEf0Ggzuw57ZzrlLqW6CCUuppYDXwg2vV8kBcmYE2IwPefBM++giWLr102tfbl1uvvdXxnVCRkXrHjvFXGAwGJ2HPbqgJwDxgPtpv8Z6IfOlqxTyOgACdpnrfPuf3HRgIJ06Avz+8+OKlaPEJPSawbMAyx/szzm2DweBk7LplFZFVIjJcRF4D/lJKDXCxXp5HZgZaV+SJmjFD78mfNw8OH4ZPPilYf8ZYGAwGJ5NXPYtySqk3lVJfKaV6KM3zwGGgX+Gp6CE0bQplyzp/vfrCBZg/X0f63nOPLrbz0Uds/edXbphyA7tO73K8z5AQqFEDqlgozmMwGAw5kNfMYhZ62SkEeApYA/QF7heR+wpBN88iMwOts43F/PmQkACDBunXEyaAry9VRowm9HSotWpsJs2HwWBwMnkZiwYi8riIfAv0B5oBd4rIjsJRzQMJCIBdu5ybgXbmTGjUCDp10q9r14ZRo6i3MYxHDpemZpmajvWXlga7dxtjYTAYnEpexiI184mIpAPhmQkFr1oyM9Bu3+6c/o4fhzVr4LHHLk8298ILHK5dignL0lCOGqaDB3UaCmMsDAaDE8nLWLRSSp23HfFAy8znSqnzecgVX5ydgXbWLB0TMXDgZafFx4cX7vWi5tnky+Iu7MI4tw0GgwvIq6yqt4iUsx1lRcQny/Nyhamkx1CjBtSt6xxjIaJ3Qd122xUpOS6mXaTSXfdztGdnGD9el0e1l5AQHbXdrFnBdTQYDAYbLssFoZT6USkVpZQKzXKuklJqlVLqgO2xYjaZG5VSaUqpPlnODbK1P6CUGuQqfe0mjwy04zeMZ27YXPv62bQJDhz4z7GdhVK+pZj1wCzqfT8X/PzghRe0cbGHkBDtA/H3t6+9wWAw2IErEwdNB+7Kdm4E8JeINAb+sr0GQCnlDXwMrMxyrhLwPhAAdADez25gCp2AADh2DE6fvux0QkoCb//9No8ueJSNJzbm38+MGfoHvU+fKy4lpSUhInomM3q0rs62YIF9+pmdUAaDwQW4zFiIyDogW5UY7gNm2J7P4PK6GC+go8Sz5gG/E1glIjEicg5YxZUGqHDJxW+x9thaUjNS8fPxo8/cPkTGR+beR1KSrqrWu7eO3cjG0KVDaT7ZljzwueegVStdyzkhIW/dEhJ0UJ8xFgaDwckUdkrS6iKS+St6CqgOoJSqjc5uOyVb+9rAiSyvw8mlloZS6hml1Fal1Nbo6Gjnap2Vtm1zzEDbpV4X/hzwJ38P+pvyfuWJTsxDh8WLITY2xyUo0KnJa5ezvU0fH/j6awgP17OMvAgL08tVxlgYDAYn47b81SIiQOZC/OfAGyJiOUufiHwnIu1FpH3VqlWdomOOlCqli95kMxalfEtxZ6M7aV+rPaFDQ2lZvWXufcyYoeMpuna94lKGZBAWHXZ5WvKbb4bHH4eJE3Vd59wwO6EMBoOLKGxjcVopVRPA9pi55NQe+EUpdRToA0xWSt0PRADXNtZXjwAAIABJREFUZJGvYzvnXgICYMuWSxloT8af5L0173EiTk+CvL28SU1P5eU/X2bGjhmXy54+rQvmPPpojlXVjsYeJTE18cqCRx9/DGXK6GpsuTm7Q0K0H6RBgwK/RYPBYMhKYRuLP4DMtZdBwCIAEakvIvVEpB46w+1zIvI7sALooZSqaHNs97Cdcy8dOlyWgXbFwRWMXjea2KTYS02UUuw6vYshS4ewPTJLEN/s2ZCenusSVGbBoyuMRbVqMHasLuc5N5cdVyEh0Ly5Ke1pMBicjiu3zs5B1+tuopQKV0o9CYwDuiulDgDdbK9zRURigNHAFtvxge2ce8nm5F55eCU1ytS47Afex8uHX/v8StVSVen9a2/OJJ7RF2bM0Dmmrr8+x64bVGzAGze/kXN1vGef1T6TV16B+Pgrr5udUAaDwUW4cjdUfxGpKSK+IlJHRKaKyFkRuUNEGotIt5x++G35qOZlef2jiDSyHdNcpa9DNG0K5crBpk1kSAarD6+mR8MeV5TprFq6KvP7zefUhVP0n9+ftOBtsHNnrrMK0DOKcd3GUbZkDpVrvb1h8mRd3GjUqMuvnT4N0dHGWBgMBpdgCjRbwcvrUgbaHad2cCbxDN0bdM+x6Y21b2TyvZPZcHwDZ7/9DHx9dRryXAiNCiUxNTH3sQMC4Kmn4PPPdQ3oTIxz22AwuBBjLKxiy0B7JCKMMiXK0K1Bt1ybPtHmCfYP3U3131dDz55QuXKO7VLTU2n3XTtGBY7K8folPvxQF0saNuw/Z7cxFgaDwYUYY2GVgABIT+fBpPrEvB5DjTI18mxeJygMTp/mr851CIsKy7HNwZiDpKSnXOnczk6VKjBuHKxbBz//rM+FhEDVqlC9upV3YzAYDHlijIVVsji5fb19828/YwYZVSrzePKvPPDrA8QlxV3RJCxaG5Hm1XJwbmfnySf1rqzXXoO4OOPcNhgMLsUYC6tUr87F2tVYOXs0B2MO5t323Dn44w+8+j/C7P7zOBJ7hIELB5KRLQYxNCoUheL6KjnvlLoMLy8d2R0VBe+8o6O3jbEwGAwuwhiLArC3YXmaHoqjdtkcM5D8x9y5uiDRoEF0vrYzn/b4lMX7FzN23djLmoVGhdKoUiP8fe3MGNu+PQwZAl99pav3GWNhMBhchDEWBWBVlfPUjQP/s1cuKV3GjBk6WK5tWwCe7/A8A1sOZOTakew7s+9Ss+GdhvPpnZ86psSYMdqHAcZYGAwGl2GMhUXCz4ezqIItTXlexZD274egIB1bYYvDUErxTc9vWNJ/CU2qNLnUNKBOAD2v6+mYIpUq6eWoli2NsTAYDC7DGAuLrD68mu01dQlUNm/OveGsWdq/MGDAZadL+Zbi7sZ3A7AlYgsHYw6yaO8izidbqFjbr58O9jMFjwwGg4vwcbcCRZW65evy8I2PQ8tduc8sMjJg5kzo3h1q1cqxSWR8JLdOv5UqpaoQfj6c0KGh9u2GMhgMhkLEzCws0rV+V6bdNw2VLQPtZaxdC8eP55neo2bZmrx/2/uEnw/H18uXxpUbu1Brg8FgsIYxFhaISojieNxx/SIzA+3evVc2nDFD55C6//4rr2XhjZvfYMANA+hSrwslvEu4QGODwWAoGMZYWGBa8DSu/fxaohKici2zyoULMG8e9O2bry9BKcWsB2ax4lH3Z183GAyGnDDGwgKrDq/ihmo3UK10NWjSROdpym4sFi7UNbHzWILKilLqiqy1BoPB4CkYY+EgiamJ/HP8H3o07KFPZMlAexkzZuiKdbfcUvhKGgwGg5MxxsJB1h1bR0p6yn/GAvRSVEgIJNpSi584oSvaPfbYpdgKg8FgKMoYY+EgKw+tpKR3STrX7fzfSVsGWrZt069nzdKpwwcOdI+SBoPB4GSMsXCQ129+ncX9F1+evymrk1tEx1Z07qyXoQwGg6EYYILyHKRGmRpX1q6oVg3q1dPGYvNm2LcPhg93i34Gg8HgCszMwgH+PvI3kzZNIiU95cqLAQHaWMyYAX5+0KdP4StoMBgMLsLMLBxgavBU/jr8F8M6DLvyYkAA/PqrXoJ64AG9ndZQZEhNTSU8PJykpCR3q2IwuBw/Pz/q1KmDr68dhdtsGGNhJxmSwapDq+jRsAdeKocJWabfwoHYCoPnEB4eTtmyZalXr56JdzEUa0SEs2fPEh4eTv369e2WM8tQdrLr9C6iE6Mv3zKblTZtwMcHataEbt0KVzlDgUlKSqJy5crGUBiKPUopKleu7PAs2sws7GTloZUAdG/QPecG/v4wbBg0awbe3oWomcFZGENhuFqw8l03xsJOjscdp1X1Vv/f3rlHR1Vdj/+zCSAPUUTQH28wSyDvIYGQIIREBELBAIUIFCmgXxRbi9rVH9D6q2K/uopCi6BYigoBhIiAPESq4RUeAkJCE8SAQCCIQEl4GAlSMcn+/TE34ySZySSQmSCcz1qz5t5z9j1n33PvnD3ntQ/NGzV3L/T6675TyGAwGHyI6YaqJG/+4k32jK9gkyOD4Trx8/PDZrMRHBzMww8/zLfffnvNabVr145z5865jMvIyEBE+OSTT0qF33777R7TrYyMM1OnTmXGjBmVlv/2229566233Ma/8sorBAUFERoais1m43PLzc7rr7/O9yUeFKpIZXScOnUqLVu2dDyftWvXupRbu3Yt06ZNuyY9bnSMsagCxn24wZvUr1+fjIwMDhw4QJMmTZgzZ45X8klOTqZHjx4kJyd7Jf3roSJjsWvXLtatW8e+ffvYv38/GzdupHXr1sD1GYvK8txzz5GRkcHy5ct57LHHKC6zh01hYSEJCQlMmTLFq3rUFMZYVILnNz1PQnICqlrTqhh8RGxSbLnPW3vtldj3P37vMj4pIwmAc9+fKxdXVaKjozl16pTjfPr06XTt2pXQ0FBefPFFR/jgwYOJiIggKCiIefPmeUxXVVm+fDlJSUls2LDB5SBnamoqMTExDBgwgI4dOzJhwoRSFePzzz9PWFgYUVFRnD1r34f+o48+olu3bnTu3JmHHnrIEQ6QmZlJdHQ0999/P2+//XaF9zRlyhSys7Ox2Wz83zILW8+cOUPTpk257bbbAGjatCktWrRg9uzZnD59mri4OOLi4gC7QQwJCSE4OJjJkyc70vjkk08IDw8nLCyM3r17l7v3t99+m/79+3PlyhW3ZRgQEEDt2rU5d+4cY8eOZcKECXTr1o1JkyaRlJTE008/DcDZs2cZMmQIYWFhhIWFsXPnTgDee+89IiMjsdlsPPnkkxQVFbnN60bCGItKsPbwWq4UXjEDoAafUFRUxKZNm0hISAAgJSWFI0eOsGfPHjIyMkhPT2fbtm0AzJ8/n/T0dNLS0pg9ezbnz5+vMO2dO3fSvn17/P39iY2N5eOPP3Ypt2fPHt544w2ysrLIzs7mww8/BODy5ctERUWRmZlJTEyMo/Lv0aMHu3fv5t///jcjRozgtddec6S1f/9+Nm/ezK5du/jLX/7C6dOn3d7TtGnT8Pf3JyMjg+nTp5fSqW/fvpw8eZIOHTrwm9/8hq1btwIwceJEWrRowZYtW9iyZQunT59m8uTJbN68mYyMDPbu3cvq1avJy8tj/PjxrFy5kszMTJYvX14q/TfffJN169axevVq6lewB83nn39OrVq1aNasGWCfdr1z507+/ve/l5KbOHEivXr1IjMzk3379hEUFMTBgwdZtmwZn332GRkZGfj5+bFkyZIKn9mNghng9sDpS6c5kHuA0aHGKeCtROrYVLdxDeo0qDC+aYOmFca748qVK9hsNk6dOkVAQAB9+thn3qWkpJCSkkLnzp0BKCgo4MiRI8TExDB79mxWrVoFwMmTJzly5Ah333232zySk5MZMWIEACNGjGDRokUMHTq0nFxkZCT3Wb7NRo4cyY4dOxg2bBh169Zl4MCBAERERLBhwwbAXmEOHz6cM2fOcPXq1VLz9wcNGkT9+vWpX78+cXFx7Nmzhx07dri8pzZt2rjV/fbbbyc9PZ3t27ezZcsWhg8fzrRp0xg7dmwpub179xIbG+uozEeNGsW2bdvw8/MjJibGoVuTJk0c1yxatIjWrVuzevVqtwvVZs6cyXvvvUejRo1YtmyZ489jYmIifi5mQG7evJlFixYB9vGoO++8k8WLF5Oenk7Xrl0B+zO/55573N7zjYQxFh7YeGwjgPv1FQZDNVEyZvH999/Tr18/5syZw8SJE1FV/vjHP/Lkk0+Wkk9NTWXjxo3s2rWLBg0aEBsbW+Hc+aKiIlauXMmaNWt45ZVXHIuzLl26RKNGjUrJlm1Fl5zXqVPHcezn50dhYSEAv/vd7/j9739PQkICqampTJ06tcK03N1TTk5OhWXk5+dHbGwssbGxhISEsHDhwnLG4loICQkhIyOjwoVqzz33HH/4wx/KhTds2LDS+agqY8aM4a9//es161pTmG4oD6Rkp3BPw3sIvTe0plUx3CI0aNCA2bNn87e//Y3CwkL69evH/PnzKSgoAODUqVPk5uaSn5/PXXfdRYMGDTh06BC7d++uMN1NmzYRGhrKyZMnycnJ4cSJEwwdOtTRMnFmz549HD9+nOLiYpYtW0YPD5t45efn07JlSwAWLlxYKm7NmjX897//5fz586SmptK1a1e399SoUSMuXbrkMo+vvvqKI0eOOM4zMjJo27YtQKnrIiMj2bp1K+fOnaOoqIjk5GR69epFVFQU27Zt4/jx4wBcuHDBkVbnzp355z//SUJCAqdPn67wXitL7969+cc//gHYDXV+fj69e/dmxYoV5ObmOnQ4ceJEteTnbYyx8EBUqygmRk507eLDYPASnTt3JjQ0lOTkZPr27cuvfvUroqOjCQkJYdiwYVy6dIn4+HgKCwsJCAhgypQpREVFVZhmcnIyQ4YMKRU2dOhQl7OiunbtytNPP01AQADt27cvd11Zpk6dSmJiIhERETRt2rRUXGhoKHFxcURFRfHnP/+ZFi1auL2nu+++mwceeIDg4OByA9wFBQWMGTOGwMBAQkNDycrKcrRgnnjiCeLj44mLi6N58+ZMmzaNuLg4wsLCiIiIYNCgQTRr1ox58+bxy1/+krCwMIYPH14q/R49ejBjxgwGDBjgdtpxVZg1axZbtmwhJCSEiIgIsrKyCAwM5OWXX6Zv376EhobSp08fzpw5c915+QK5GWf4dOnSRdPS0mpaDcPPiIMHDxIQEFDTatwQpKamMmPGDNatW1fTqhi8iKt3XkTSVbWLK3nzd7kCsi9k890P39W0GgaDwVDjmAHuCpjw8QRyL+eSOSGzplUxGHxGyQCyweCMaVm44cqPV9h+Yju925dfuGMwGAy3Gl4zFiIyX0RyReSAU1gTEdkgIkes77us8FEisl9EvhCRnSIS5nRNvIh8JSJHRcRn6+i3f72dH4p+MFNmDQaDAe+2LJKA+DJhU4BNqno/sMk6BzgO9FLVEOB/gXkAIuIHzAH6A4HASBEJ9KLODlKyU6jrV5eYtjG+yM5gMBhuaLxmLFR1G3ChTPAgoGQS9kJgsCW7U1UvWuG7gVbWcSRwVFWPqepV4H0rDa+Tkp1CzzY9aVCngS+yMxgMhhsaX49Z3KuqJZOK/wPc60LmceBf1nFL4KRT3DdWWDlE5AkRSRORtLy8vOtWdOnQpfy1989vlaXh54uzi/LExMTr8qKamprqcMvhyW22J7fg7vDk2ttmszlci5QwduxYVqxYUWG6lZFxJicnh+Dg4ErLAyQlJbldfLd79266deuGzWYjICDAsZYjNTXV4QywqlRGx5ycHOrXr4/NZiMwMLCcA0dnunfvfk16XA81NsCt9gUepRZ5iEgcdmMx2eVFFac3T1W7qGqXEp8w10PwPcF0bdn1utMxGCqLs4vyunXrMnfu3FLxquq28qgIT26zr9VYVMTBgwcpKipi+/btXL58uVrTrg4qMhZjxoxh3rx5jmfxyCOPANdnLCpLiRPF/fv3k5WVxerVq0vFl7hX8bYervC1sTgrIs0BrO/ckggRCQXeAQapaonrzFNAa6frW1lhXuWdfe/w0VcfeTsbw43Ks89CbGz1fp59tkoq9OzZk6NHj5KTk0PHjh359a9/TXBwMCdPniQlJYXo6GjCw8NJTEx0uMz45JNP6NSpE+Hh4Q4vsYBHt9mu3IK7c4n+yiuv0KFDB3r06MFXX33lVv/k5GRGjx5N3759WbNmjUuZdu3aMWnSJEJCQoiMjOTo0aOOuG3bttG9e3fuu+8+RyujoKCA3r17Ex4eTkhISKl0CwsLGTVqFAEBAQwbNszRKktPT6dXr15ERETQr18/zpw5w4oVK0hLS2PUqFHYbLZy7shzc3Np3ty+I6afnx+BgYHk5OQwd+5cZs6cic1mY/v27eTk5PDggw8SGhpK7969+frrr92WsTPHjh2jc+fO7N2712351a5dm+7du3P06FFSU1Pp2bMnCQkJBAbah2ydN6F69dVXCQkJISwszPGnIDs7m/j4eCIiIujZsyeHDh1ym1elUVWvfYB2wAGn8+nAFOt4CvCaddwGOAp0L3N9beAY0B6oC2QCQZ7yjYiI0GuluLhY751+r45cMfKa0zD8/MjKyvrp5JlnVHv1qt7PM8941KFhw4aqqvrjjz9qQkKCvvXWW3r8+HEVEd21a5eqqubl5WnPnj21oKBAVVWnTZumL730kl65ckVbtWqlhw8f1uLiYk1MTNQBAwaoquqCBQv0t7/9raqqPvLIIzpz5kxVVS0sLNRvv/1Wjx8/rkFBQQ49Pv30Ux0/frwWFxdrUVGRDhgwQLdu3appaWkaHBysly9f1vz8fPX399fp06e7vJcOHTroiRMn9NNPP9WBAwc6wseMGaPLly9XVdW2bdvqyy+/rKqqCxcudOg7ZswYHTZsmBYVFemXX36p/v7+jnLJz893lIO/v78WFxfr8ePHFdAdO3aoquq4ceN0+vTpevXqVY2Ojtbc3FxVVX3//fd13Lhxqqraq1cv3bt3r0vdX3rpJW3cuLEOHjxY586dq1euXFFV1RdffLHU/Q4cOFCTkpJUVfXdd9/VQYMGeSzjQ4cOqc1m04yMjHL5Oj+Hy5cva5cuXXT9+vW6ZcsWbdCggR47dswhW/KurF+/XqOjo/Xy5cuqqnr+/HlVVX3wwQf18OHDqqq6e/dujYuLK5dfqXfeAkhTN/Wq1xbliUgyEAs0FZFvgBeBacAHIvI4cAJ4xBJ/AbgbeMvyUFmo9i6lQhF5GvgU8APmq+qX3tIZ4IvcLzh7+ayZMnsrU0N7qZe4KAd7y+Lxxx/n9OnTtG3b1uH3affu3WRlZfHAAw8AcPXqVaKjozl06BDt27fn/vvvB+DRRx91uRmSK7fZFy9eLCXjziX6pUuXGDJkCA0a2Cd9lOy3UZa0tDSaNm1KmzZtaNmyJY899hgXLlwo5RK8hJEjRzq+n3vuOUf44MGDqVWrFoGBgY6NlFSVP/3pT2zbto1atWpx6tQpR1zr1q0dZfLoo48ye/Zs4uPjOXDggMPVe1FRkaPFUBEvvPACo0aNIiUlhaVLl5KcnExqamo5uV27djlacKNHj2bSpEmA+zLOy8tj0KBBfPjhh44WQllKWngiwqBBg+jfvz+pqalERka69Ia7ceNGxo0b53gmTZo0oaCggJ07d5KYmOiQ++GHHzzetye8ZixUdaSbqHKr3FT1f4D/cZPOemB9NapWISnZKQD0ua+Pr7I0GICfxizK4uwCW1Xp06dPOed/rq67VtSN+/DXK2lEk5OTOXToEO3atQPgu+++Y+XKlYwfP76crLP7cufjkt3wSvQBWLJkCXl5eaSnp1OnTh3atWvncMnuzg16UFAQu3btqpTezvj7+/PUU08xfvx4mjVr5nFTqcpw55130qZNG3bs2OHWWJSMWZSlKm7Qi4uLady4cbW+E2BWcJcjJTuFoGZBtLzD5aQrg6FGiYqK4rPPPnP071++fJnDhw/TqVMncnJyyM7OBnC7v7Yrt9ll3YK7cx8eExPD6tWruXLlCpcuXeKjj8qP6xUXF/PBBx/wxRdfkJOTQ05ODmvWrHGrz7Jlyxzf0dHRFd57fn4+99xzD3Xq1GHLli2lXHt//fXXDqOwdOlSevToQceOHcnLy3OE//jjj3z5pb1joiJX6B9//LHDQB05cgQ/Pz8aN25c7pru3bvz/vvvA3ZD1rNnT7dlDFC3bl1WrVrFokWLWLp0aYX3Wln69OnDggULHGM0Fy5c4I477qB9+/aOnQBVlczM63dZZIyFE8VazDfffWO6oAw3LM2aNSMpKYmRI0cSGhrq6IKqV68e8+bNY8CAAYSHh7vdfc2V2+yybsHduQ8PDw9n+PDhhIWF0b9/f8dub85s376dli1b0qJFC0dYTEwMWVlZLl1xX7x4kdDQUGbNmsXMmTMrvPdRo0aRlpZGSEgIixYtolOnTo64jh07MmfOHAICArh48SJPPfUUdevWZcWKFUyePJmwsDBsNptjsLlk72xXA9yLFy+mY8eO2Gw2Ro8ezZIlS/Dz8+Phhx9m1apVjgHuN954gwULFhAaGsrixYuZNWuW2zIuoWHDhqxbt46ZM2eydu3aCu+3MsTHx5OQkECXLl2w2WyOqcxLlizh3XffJSwsjKCgILeTDKqCcVFeBlXlatFVbqt9m2dhw02DcVHue9q1a+cY3zD4HuOi/DoREWMoDAaDoQzGRbnBYKgRPO23bbixMC0Lg8HiZuySNRhccS3vujEWBgNQr149zp8/bwyG4aZHVTl//jz16tWr0nWmG8pgAFq1asU333xDdTihNBhudOrVq0erVq08CzphjIXBANSpU8flClmDwWDHdEMZDAaDwSPGWBgMBoPBI8ZYGAwGg8EjN+UKbhHJw+7V9lppCpyrJnWqE6NX1TB6VQ2jV9W4GfVqq6oud4+7KY3F9SIiae6WvNckRq+qYfSqGkavqnGr6WW6oQwGg8HgEWMsDAaDweARYyxcU36LsRsDo1fVMHpVDaNX1bil9DJjFgaDwWDwiGlZGAwGg8EjxlgYDAaDwSO3rLEQkXgR+UpEjorIFBfxt4nIMiv+cxFp5wOdWovIFhHJEpEvReQZFzKxIpIvIhnW5wVv6+WUd46IfGHlW24rQrEz2yqz/SIS7gOdOjqVRYaIfCciz5aR8UmZich8EckVkQNOYU1EZIOIHLG+73Jz7RhL5oiIjPGBXtNF5JD1nFaJSGM311b4zL2g11QROeX0rH7h5toKf79e0GuZk045IpLh5lpvlpfL+sFn75iq3nIfwA/IBu4D6gKZQGAZmd8Ac63jEcAyH+jVHAi3jhsBh13oFQusq6FyywGaVhD/C+BfgABRwOc18Fz/g31hkc/LDIgBwoEDTmGvAVOs4ynAqy6uawIcs77vso7v8rJefYHa1vGrrvSqzDP3gl5TgT9U4jlX+Putbr3KxP8NeKEGystl/eCrd+xWbVlEAkdV9ZiqXgXeBwaVkRkELLSOVwC9RUS8qZSqnlHVfdbxJeAg0NKbeVYzg4BFamc30FhEmvsw/95Atqpez+r9a0ZVtwEXygQ7v0cLgcEuLu0HbFDVC6p6EdgAxHtTL1VNUdVC63Q3UDV/1V7Sq5JU5vfrFb2sOuARILm68qssFdQPPnnHblVj0RI46XT+DeUrZYeM9aPKB+72iXaA1e3VGfjcRXS0iGSKyL9EJMhXOgEKpIhIuog84SK+MuXqTUbg/kdcU2V2r6qesY7/A9zrQqamy+0x7C1CV3h65t7gaat7bL6bLpWaLK+ewFlVPeIm3iflVaZ+8Mk7dqsaixsaEbkdWAk8q6rflYneh72bJQx4A1jtQ9V6qGo40B/4rYjE+DDvChGRukACsNxFdE2WmQO19wfcUHPVReR5oBBY4kbE18/8H4A/YAPOYO/yuZEYScWtCq+XV0X1gzffsVvVWJwCWjudt7LCXMqISG3gTuC8txUTkTrYX4Qlqvph2XhV/U5VC6zj9UAdEWnqbb2s/E5Z37nAKuzdAc5Uply9RX9gn6qeLRtRk2UGnC3pirO+c13I1Ei5ichYYCAwyqpkylGJZ16tqOpZVS1S1WLgbTf51VR51QZ+CSxzJ+Pt8nJTP/jkHbtVjcVe4H4RaW/9Ix0BrC0jsxYomTEwDNjs7gdVXVj9oe8CB1X1725k/k/J2ImIRGJ/hr4wYg1FpFHJMfYB0gNlxNYCvxY7UUC+U/PY27j9x1dTZWbh/B6NAda4kPkU6Csid1ndLn2tMK8hIvHAJCBBVb93I1OZZ17dejmPcQ1xk19lfr/e4CHgkKp+4yrS2+VVQf3gm3fMG6P2P4cP9pk7h7HPqnjeCvsL9h8PQD3sXRpHgT3AfT7QqQf2JuR+IMP6/AKYAEywZJ4GvsQ+A2Q30N1H5XWflWemlX9JmTnrJsAcq0y/ALr4SLeG2Cv/O53CfF5m2I3VGeBH7H3Cj2Mf59oEHAE2Ak0s2S7AO07XPma9a0eBcT7Q6yj2PuyS96xk5l8LYH1Fz9zLei223p392CvB5mX1ss7L/X69qZcVnlTyTjnJ+rK83NUPPnnHjLsPg8FgMHjkVu2GMhgMBkMVMMbCYDAYDB4xxsJgMBgMHjHGwmAwGAweMcbCYDAYDB4xxsJwyyEidzt5EP1PGS+nO72QXxcRmX0d148VkTerUyeDoarUrmkFDAZfo6rnsbuTQESmAgWqOsOL+aUB1equ2mDwNaZlYTA4ISIF1nesiGwVkTUickxEponIKBHZY+1X4G/JNRORlSKy1/o84CLNWBFZZx1PtRzkpVrpTnSjxzgROSwie4AHnMIfFvv+Kv8WkY0icq+I1LL2KGhmydQS+z4PzUQkUUQOWE4Ut3mhyAy3CMZYGAzuCcO+EjwAGA10UNVI4B3gd5bMLGCmqnYFhlpxnuiE3WV0JPCi5e/HgeXy4iXsRqIH9j0LSthPwh+OAAABp0lEQVQBRKlqZ+yuuSep3Y/Se8AoS+YhIFNV84AXgH5qd6KYUIV7NxhKYbqhDAb37FXLt5WIZAMpVvgXQJx1/BAQKD9tdXKHiNyuluNCN3ysqj8AP4hILnaX0s7+hroBqVZlj4gsAzpYca2AZZZBqQsct8LnY/cJ9Dp2tw4LrPDPgCQR+QAo55jSYKgspmVhMLjnB6fjYqfzYn76o1UL+z99m/Vp6cFQlE23iKr9aXsDeFNVQ4AnsfswQ1VPYvc++iD2Fsu/rPAJwP/D7nE0XUR8tieL4ebCGAuD4fpI4acuKUTEVg1pfg70smZt1QESneLu5CfX0mX3UX4He3fUclUtsvTxV9XPVfUFII/SbqoNhkpjjIXBcH1MBLqIfWe3LOxjHNeF1fU1FdiFvRvpoFP0VGC5iKQD58pcuha4nZ+6oACmWwPyB4Cd2D2iGgxVxnidNRhuEkSkC/bB9p41rYvh5sMMcBsMNwEiMgV4ip9mRBkM1YppWRgMBoPBI2bMwmAwGAweMcbCYDAYDB4xxsJgMBgMHjHGwmAwGAweMcbCYDAYDB75/4VDwONPmSiWAAAAAElFTkSuQmCC\n", + "image/png": "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\n", "text/plain": [ "
" ] diff --git a/Chapter09/Exercise9.02/Exercise9_02.ipynb b/Chapter09/Exercise9.02/Exercise9_02.ipynb old mode 100644 new mode 100755 index fb94e655..dfb94f32 --- a/Chapter09/Exercise9.02/Exercise9_02.ipynb +++ b/Chapter09/Exercise9.02/Exercise9_02.ipynb @@ -16,7 +16,8 @@ "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", - "import pandas as pd" + "import pandas as pd\n", + "from tensorflow import random" ] }, { @@ -131,7 +132,7 @@ } ], "source": [ - "dataset_training = pd.read_csv('GOOG_train.csv')\n", + "dataset_training = pd.read_csv('../GOOG_train.csv')\n", "dataset_training.head()" ] }, @@ -315,7 +316,9 @@ "metadata": {}, "outputs": [], "source": [ - "np.random.seed(1)\n", + "seed = 1\n", + "np.random.seed(seed)\n", + "random.set_seed(seed)\n", "\n", "model = Sequential()" ] @@ -615,7 +618,7 @@ } ], "source": [ - "dataset_testing = pd.read_csv(\"GOOG_test.csv\")\n", + "dataset_testing = pd.read_csv(\"../GOOG_test.csv\")\n", "actual_stock_price = dataset_testing[['Open']].values\n", "actual_stock_price" ]