{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# !pip install --upgrade git+https://github.com/stanfordmlgroup/ngboost.git","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.419314Z","iopub.execute_input":"2022-07-24T10:47:46.419843Z","iopub.status.idle":"2022-07-24T10:47:46.424256Z","shell.execute_reply.started":"2022-07-24T10:47:46.419799Z","shell.execute_reply":"2022-07-24T10:47:46.423240Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport re\npd.options.display.max_columns = 500\npd.options.display.max_rows = 500\n\n# Model Packages\nimport sklearn\n# import ngboost\nimport lightgbm as lgb\nimport xgboost as xgb\nfrom sklearn.model_selection import train_test_split, KFold, cross_val_score, GridSearchCV\nfrom sklearn.metrics import make_scorer, classification_report\nfrom catboost import CatBoostClassifier, Pool, EShapCalcType, EFeaturesSelectionAlgorithm\n\n# Plotting\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nimport plotly.express as px\nimport plotly.graph_objs as go\n%matplotlib inline\ncm = sns.light_palette(\"green\", as_cmap=True)\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-07-24T10:47:46.488757Z","iopub.execute_input":"2022-07-24T10:47:46.489061Z","iopub.status.idle":"2022-07-24T10:47:46.507337Z","shell.execute_reply.started":"2022-07-24T10:47:46.489029Z","shell.execute_reply":"2022-07-24T10:47:46.506230Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Load Dataset","metadata":{}},{"cell_type":"code","source":"df_train = pd.read_csv(\"/kaggle/input/titanic/train.csv\")\nprint(df_train.shape)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.559409Z","iopub.execute_input":"2022-07-24T10:47:46.560804Z","iopub.status.idle":"2022-07-24T10:47:46.572779Z","shell.execute_reply.started":"2022-07-24T10:47:46.560754Z","shell.execute_reply":"2022-07-24T10:47:46.571666Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_test = pd.read_csv(\"/kaggle/input/titanic/test.csv\")\nPassengerId = df_test['PassengerId']\nprint(df_test.shape)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.621747Z","iopub.execute_input":"2022-07-24T10:47:46.622853Z","iopub.status.idle":"2022-07-24T10:47:46.633977Z","shell.execute_reply.started":"2022-07-24T10:47:46.622784Z","shell.execute_reply":"2022-07-24T10:47:46.632917Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Explore Data\n\nI recommend [this notebook](https://github.com/AnelMusic/Kaggle_Titanic_Advanced_DataAnalysis/blob/main/Titanic_DataAnalysis_Top3.ipynb) for those who would like to get more infomation from plots.","metadata":{}},{"cell_type":"code","source":"display(df_train.head())\ndisplay(df_test.head())","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.700690Z","iopub.execute_input":"2022-07-24T10:47:46.701374Z","iopub.status.idle":"2022-07-24T10:47:46.733323Z","shell.execute_reply.started":"2022-07-24T10:47:46.701327Z","shell.execute_reply":"2022-07-24T10:47:46.732411Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(df_train.columns)\nprint(df_test.columns)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.748890Z","iopub.execute_input":"2022-07-24T10:47:46.749633Z","iopub.status.idle":"2022-07-24T10:47:46.757406Z","shell.execute_reply.started":"2022-07-24T10:47:46.749589Z","shell.execute_reply":"2022-07-24T10:47:46.755998Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"display(df_train.isnull().mean().T)\ndisplay(df_test.isnull().mean().T)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.851237Z","iopub.execute_input":"2022-07-24T10:47:46.851520Z","iopub.status.idle":"2022-07-24T10:47:46.870152Z","shell.execute_reply.started":"2022-07-24T10:47:46.851476Z","shell.execute_reply":"2022-07-24T10:47:46.869161Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.Ticket.nunique()\n# df_train.Ticket.unique()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.917387Z","iopub.execute_input":"2022-07-24T10:47:46.917709Z","iopub.status.idle":"2022-07-24T10:47:46.925593Z","shell.execute_reply.started":"2022-07-24T10:47:46.917674Z","shell.execute_reply":"2022-07-24T10:47:46.924594Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Fare'] = round(df_train['Fare'])\nprint(df_train.Fare.quantile([0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]))\nsns.displot(df_train[df_train.Fare<50].Fare)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:46.985523Z","iopub.execute_input":"2022-07-24T10:47:46.985828Z","iopub.status.idle":"2022-07-24T10:47:47.254120Z","shell.execute_reply.started":"2022-07-24T10:47:46.985797Z","shell.execute_reply":"2022-07-24T10:47:47.253209Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.Age.describe()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.256021Z","iopub.execute_input":"2022-07-24T10:47:47.256348Z","iopub.status.idle":"2022-07-24T10:47:47.268223Z","shell.execute_reply.started":"2022-07-24T10:47:47.256307Z","shell.execute_reply":"2022-07-24T10:47:47.267207Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(df_test.Name.nunique())\ndf_test.head(10).Name.unique()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.269523Z","iopub.execute_input":"2022-07-24T10:47:47.269808Z","iopub.status.idle":"2022-07-24T10:47:47.280822Z","shell.execute_reply.started":"2022-07-24T10:47:47.269774Z","shell.execute_reply":"2022-07-24T10:47:47.279712Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.Embarked.unique()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.283026Z","iopub.execute_input":"2022-07-24T10:47:47.283666Z","iopub.status.idle":"2022-07-24T10:47:47.292337Z","shell.execute_reply.started":"2022-07-24T10:47:47.283525Z","shell.execute_reply":"2022-07-24T10:47:47.291573Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.Cabin.unique()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.293441Z","iopub.execute_input":"2022-07-24T10:47:47.294211Z","iopub.status.idle":"2022-07-24T10:47:47.307401Z","shell.execute_reply.started":"2022-07-24T10:47:47.294163Z","shell.execute_reply":"2022-07-24T10:47:47.306701Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.Sex.unique()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.309001Z","iopub.execute_input":"2022-07-24T10:47:47.309385Z","iopub.status.idle":"2022-07-24T10:47:47.322360Z","shell.execute_reply.started":"2022-07-24T10:47:47.309344Z","shell.execute_reply":"2022-07-24T10:47:47.321363Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.Fare.unique()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.365470Z","iopub.execute_input":"2022-07-24T10:47:47.365788Z","iopub.status.idle":"2022-07-24T10:47:47.371884Z","shell.execute_reply.started":"2022-07-24T10:47:47.365756Z","shell.execute_reply":"2022-07-24T10:47:47.371299Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def plot_distribution(df, feature, title, bins=30, hist = True, fsize = (5,5)):\n    fig, ax = plt.subplots(figsize=fsize)\n    ax.set_title(title)\n    sns.distplot(df[feature], color='orange', bins=bins, ax=ax)\n    fig.legend()\n    return\n\ndef plot_kde_survivors(df, feature, title, fsize = (5,5)):\n    fig, ax = plt.subplots(figsize=fsize)\n    ax.set_title(title) \n    sns.kdeplot(df[feature].loc[df[\"Survived\"] == 1],\n                shade= True, ax=ax, label='Survived').set_xlabel(feature)\n    sns.kdeplot(df[feature].loc[df[\"Survived\"] == 0],\n                shade=True, ax=ax, label=\"Not Survived\")\n    fig.legend()\n    return\n\nplot_distribution(df_train, 'Age', 'Training Age Distribution')\nplot_distribution(df_test, 'Age', 'Testing Age Distribution')","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.428509Z","iopub.execute_input":"2022-07-24T10:47:47.428806Z","iopub.status.idle":"2022-07-24T10:47:47.896295Z","shell.execute_reply.started":"2022-07-24T10:47:47.428772Z","shell.execute_reply":"2022-07-24T10:47:47.895462Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Feature Engineering\n\nI followed this regular expression tutorial to get the string in name.\n\n[Chinese Version Tutorial](https://blog.techbridge.cc/2020/05/14/introduction-to-regular-expression/)\n\n[English Version Tutorial](https://refrf.dev/)\n\n[Python Library](https://docs.python.org/3/library/re.html)\n\nI also referred to this [Kaggle notebook](https://www.kaggle.com/code/gauravduttakiit/regular-expressions/notebook) for the following two tables:\n\n![image.png](attachment:4c26e2ca-1a22-4259-b7a9-5c599204a137.png)\n![image.png](attachment:148df854-4102-44ad-aa3d-44cee56dc924.png)\n\nFor the feature analysis before engineering, it's interesting to see [this article](https://www.shiftcomm.com/insights/never-let-go-titanic-survival-101/).","metadata":{},"attachments":{"4c26e2ca-1a22-4259-b7a9-5c599204a137.png":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAawAAADXCAYAAABGWmUaAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAADsMAAA7DAcdvqGQAAF8DSURBVHhe7Z0HXBTHF8d/dyhBzwIiViyACohgwd5NtYEYjWJBjS0xUWPXWGISuzGJxpJYYvtbsHeNxoYooIgiSJOqohRFEDnFQ9j/7u1egwOOO4yevu/nsx9uZmdmZ2fevDdvdpYVMSwgCIIgiLccsfCXIAiCIN5qlB5WfEICypQpI48kXi+vXr2itiZeKyRjxLuCuiwrDdb9xERYVqkijyReL2lPnlBbE68VkjHiXUFdlmlJkCAIgjAKyGARBEEQRgEZLIIgCMIoIINFEARBGAVksAiCIAijgAwWQRAEYRSQwSIIgiCMAjJYRKkhk0qRI/w2hNIqhyCIdwuDXhxmcjMQcmI3dp85j5hkBmJza3T5dCD69mqHWhIhDaKxZ+AR1F87FW2rivjItwxOQb4qK0F5UyHiNVPSlzpzn1zBL1+vxPU8/t8+mtRoiMbV6qBpt4/Qrb09KpvIo98ouffPYPaUjbCdtAFfdbAQYosnf9vrWw6hSclkLAPnV36NdZdfCWEVFgMWYuMAByH0dvIq5hBGHbHE6qmdUUmIKylh+wdjvrdMCAE1GjaBnW0bdO/fHY4WJdBbuTJIswGJRFOZ/Nc6pjCe3b+M/bsO4HLgfTxlw7WafQb3Pv3xsbPhY43rB6+Zr7Bo/xewLUVVXyovDjPZ0di7cAbWh1mh39S12LZtG/76aThqPtiJ6Qv2I/aFkPCtJwOXN47ClqvpQvjtRMS0xsRft8rb+e+F38LTzRHp53/FN3PejrY2qdMBUxYvQj/Xkgh+wbbXrxyiNOjy1e9y+VI/Vrm/3caqNKnc93vlfS8YPxitq97C79NXIeix7v8f/FX8CYwedgRxGlneDh3z0P8vzJh8AuJWI/HzJu4+12FcRzNc+mM61vumCKnebvQ2WNHntuNkhaH4+fv+aGVTjZ1RSFCldnP0nTwfo2r/i+2n44SUanCzD3amIZWqZjLqMDLunPbz3AxFlsunSYlPxHMhnqOofAqYbD7Nc3b2o85LNi6HLTcvO0t+Pv9SFHddeb78RcvvhY98mZGIu48Kv3ZpYcq2MdfOFVhP1r55N4yYv0je1n/sCy24hFZMW7N3Jpzn21Ubhd279r6QwKJWbVQ2kwfk5xRtrSgn/3W0t71mOSpU9c1/r7pciygesVkFuXypH+XV+rOgLPF9kr+tC5Obgv2klqAweVUbZ8oxnL8ahVDYmC8MMesBKe7bonYjdOg7AV6uATjqHy+k4OHbomBd5fFZMjBsu2SpXbe0dYw+8p737CaObo7FJwt+xPAPnVHbnLvPanDs6oVZM3shYdV+BGVqGmZd2k+n6xeri3TH5EcW7kdmZibKlysnjywObpnv7IrTaP7leDSz1PT9RKzCqVHdAi+filGjQXV8gCesux2FD5yeYeeyP3H5VgAObtiA86l14Nq6DsrLc8kQfWoV5s7dg+txt+B3yht7/dJR17kpqpfnys+Az7qxuPGqOk4v+R4bj6bD9sP2sC6fU0w+rq6PELjtV8xfegg34wJx5M/dCM2zgWuT6iiDBPyzbAtOx6Tgwf1E3L6ZhKoubN5yIrazEnBi9Vz8uPcq7gSfhvefR/GoqgucbSqDW4F7FXcUY3feh8XDbZi1eAcCJc3h7lRVfs3iePHihc5tzcG8uA+/4/dRvzd3z6r25tq6bv2KuLoyBFY9W6C6KX8uKWgHfvlxLU6H3saVo3/hwNVXaODaBFU/4M+/SDiD3+bNwYGAaARfPYm9B2Nh1bwprCvy/2BSce/zt19E9O3z+e5de1/ULh+D/V478KJbO3kdHwesxbTrprCL+JMtxx8hfkewcc81iOu1QuMapmy/aG/7auU0y+FQr++Nf/dh09EoWDi2hK0FX1/uWlOu5aDKpRX47fhNBJzYiK2nHqF+q1aoJVG11/tEyWQsG/EBx/G4Vm+0rqc9D/M8DFsmr0KSQxfYW/Lt/uDcCkw5boKuHerJx3Hu45vY9st8bDgVitCbF3Fw3QHEV3JEiwYW8jGj6qdVWP2vP05s5PWAY5VgrP5pAy7fPI89Gw4hwrQJ2jjwebhx9tXOJ3BgjmL2bycRqmUs5j2JxNGo8ujZvh6rb1Rj/se/TuB2OKtv1MZ8Wfa8Nh6FH0AA0yXfGDZFXvotnMxyEuKL1lNpwXuxat8VpD65h2T2fHB2XXRo9LTUdYw+8v74ujc2Z3fFeHc7eRupU7ZKTVhVfIrcivVRy7yM1vYLkdaGY/Nagr7mdMQD/LthNuZuYXUEm+bQgUhY1jXF1atifDTACYpV1OJ0kS5oyDL3DIvj3v37DGsFdToy755h5vdcxwQ9ydJ6Xv3IkgYzm3r3Yr5cfIq5m8nHcfkXDB7M/C/giTz86NYeZtLElYxPvKK8dObyei9mzJYbQvgBc2yRG+MxYRsTkqK6ZvH5pMzNvVPYax9lotP48LPUYGbzpN7M6nMpQhq+7OX/JCrzSKWpzPlVQ5mZbDmPhDi+zt8w3jf5Oj+9tYPp2WMQ8+vJOLV8uh0laWvuyLx/hvmh54/MubsF21vRvntu8ueehu1hvhu0iDmnbJNU5tq2KUz/n08xSfL0CczBub2ZZWr1jj7zKzP797Py84p7n7EhoMC9HwjlytTeF1w9/u79k7KOCWcWMr0GDWf+ULtO0o0dzESN+yjY9vnLybx/nr32d4z3dUV/SZmHV9Yz3w5ayfgnqV2rx0S1NKnMlS0TmIG/+yjv4X07SiZj2sZAwePu2WXMiO+PMnfZ31y/LBzM9pNSzh4w//wykvl5ZyiTIaTPjDvBfK/W33w//cicFvI8S73M/DGkFyubbJmCbnh0YzPzTQ+2bwXZ4sZZr0GDmLnFjEWFfHNhfsyr9I1izBd1f9e29WEGbwnKF5/OnF/jzkzzDpWHddE3XF3cu29lQrPUyyldHaOPvF/b2oeZf1gXXZXO3Nyh2X5ZaXeYPQt7M7OEduDSXNs2QWsbq997cbqIjyv+UJdlvZYEmYxU3MwrD1NhRl8cIlRG956dUJWbNrCYVG2Fbl2ykJTBLyaVb9Abi5aMQctqivJMUb2ePdLvpyBTiOHo3q877CqorllcvjxZGHy3ZaO324eoJSxtiCUNMWDeZng2q8BHaCE3JRCnArpgQJ9GyhmFSdUOGOhZCedDE4UY9mr13dCzc3Uh9GYQwQrV2ym6UYbQyzthNXgQ2irbRAJHd090uXoT0emcyy9FVgZQubywK4alVoevMWdsO/kDa+7eT552Ydusica9j1/5Izpbq9o+f19oo6z5Z/hErX0q2X+Knh8G4dLNVCGmeJJvnkFIUzd85Kjqr8rN+sLN9Rx8b7E3IlC+kxu6KNNI0KRFG2RfTUAJHj+891xYNQb9+/fXOI7fUTVg1XZD4Wm+F96+9xFycgswwEtNzszR6ZvVmOphq/RixNVro644EI+fCBEsFT7thhZCHrHEBjaNxWjduoVSN5SzbwFXk6cay1Ci9Dbo1VNzLPZ2M8HhgKgCy2vcmL+8rRwGDlDpG27Mu3m6IfBGhIY+yU9eDr+8xR+pCD+1GdtOtECPdjby87rqKV0wVMeUTN4zkPIgDxKz4nd85Mmi4bub05mq9hOZ1Ub3vl54duSG/Nkcl+bqXrMCbfxZnw/5gBxddFHJ0ctgicyrobn4OWQyXS9qj6oaz9BNYWqmqeyy06Lhc9IbG377CT/+OAnzVl0RzqiwqKxSsgqKysewmjmNaY5GNprXEpmp1ua1wTxLRziuwXsZW+bPPyqP9Sfu4UnCA6VwlqlpAXPNov9zGFYYU8PzhNBzpLMSG/vvnxr1/nnpTlxjfJH0iDNw9dFjjCfubPgS47//BZv3nMOtxCwhP3/vCa2cUa+S5o1xz8/Udzhp64v8mLnYoKZGMRLYNnJihVX3h88Z6RFo6WyrHNQ8ElSvYY7gJNWD4rI1qxTYIcZkMCi4740oDG2bLj61U+9Ac7T3HIas1VPwe8IX8PoknyLNleJe8AUc/h8vf9+Pn42juQrZ5DExr5CvL4vnA1aOauWbHFnbNMGrRxnIv9+IG/OPmET8s1Fz7C7b5Y+Xlx4UOYF5emgJhg8fLj9GDJ+NzcHl4LVyosbuZl30lC4YqmNKJu/mqF5bDGl28c+QCtOZYgsr1Hgaj9Q0Rp7mMdMItWtopqlQpZrwi6N4XaQP+hmsSuaoaREGtl21kvfsAcIiUzU2RhTF0+DNmDJzB6KYeujgNgqTpy/HgskdhLOFo1u+nBIYVhVlq7fGJ5+zs0y1Y8DI7zCjjzN0f/r0+slLSUBMenvUtBIiWJp27KdR7/6fD8b4+T+ihWA9yjfoi5/XbsXMkZ1RV5SAo8u+xM97VbNV5oXhD0c5GHaA5J8BS7Olwi/d0fUhO2EY2jZdmAozaAUvn6Yio3JllM9IwVM1a8E9kzy2aDJ+O5eGqk06YcS46Zj322L0M9FLxWiQ+0yK5/mGcI6scKEQwQldNeS/P74YMg5zf2qHfDpWA24L//79++XHvv2bsOL74WhnrXrio6+eKoz/UseYWzVFeFRcITpZhuTo27jLGiMeXXSmbnq1OF1UUvSSJrFpQ7T4VIzjx3zxWIhTIcOtIwuwISil0Aec+UmMOYv6A7/B2F7t4NTQGpW5dxg0J2ZaKS6fmLX4dbUY1mcPwhD+QOVV5EdU0QKOd6SoZOeEJk2aKA+nBo3Yw0Ln+3r9SBH0rzdSP+kkvCtSHhbVzfHExFyj3k2aNEKjRk1QqyKfizMkIlMJ6jRsg48HjMLs78cgd98lRGQxELFt1uR2FO5q7BiS4UkMJ9Alsxwv7+Sf0WYgMToBzWrqvoxarWarAgONeygcH5mOzjbWQgzxX8AZpVObT6H+iPmY0DoEO4+pJjl5Tx4gIrQDvh7fDx2bN0F9K94jzz9h0YdX8Yl49ExTOSbGX4NFneoFvAxRBW4y/Yj15PKN3UbsGGhiXWLvTh199ZQ2/msdU8OlIxx9juHcnZdCjIrcFD9smn0E99jhXZjOzE5MQHS9RqhlKRLauGCadHbyrEI3XVRS9Jz+mKJx71HolLoWi9acQswjKWttZcjKSMCFbUvx50UHDPusic6NXrZcXdwLj8JjYWtkdspVHD3mzweKoLh8IrCGtc8H+N+GA4gUFHDe05vYt2olIp8oasc2LOvyJz9IRIaU33JqUr0VPvnsOg7sCsBjQUdz+Xb8PBRrfd/cuxSKLaTckRx/FQd/n4lVtz/Gd16uwkA0hXO73sjYvx2n7igMshQJ59Zj4uz9iGVlNU96E+snzIF3oOI+ZEi5E4ZYO2tYSUQwqdIMXT4LwZYt7GREuPfM+xex8fujrECXbBi9yj6NvfuiBGMjw0Pf/8H79kfo1NRcHqOt7fNj6fopOodsxnafFOG8FHfPeePw0y/QxaXw55BEyVFsu9Y8VJOUh+f24GiVoejfoQ4c3b1gfXYb/onmFaDog/KoJIpCdJzgQedKEXH0QIElQX1gWOW4Zzsrj8I4fxp2CP/bUwXdWzbiI9QQmzqhTZ9c7P/fUcQKY56RPsC/6ydggdoqgj7ooqc4Q9RQnIAH7IRY9RzuzesYkyod4DWpJo4uX4z9AffxlO1XWbYUSWHHsGLhWjCeg9CW9Xo4ndlpoGUBnXlg5zk068utyCjamNersewklyM75TK8D8ZAZYeK10X6oLe/LjJriAGzf0Mf82Cs+GY4Bg8ejC9HL8TJJ00x87fv4FqC/2rR8KOv0T1vL74eyLqM/ftj8rr7cPyorXC2cHTJ16jXLMxpfw8rR30hTzNw1EZIu3wPN2eFq28Kl+6TUC3wD4wZPh/n5G/8SdB66A/oUuZfzBzMl83lS2+9AF92ejMvtDKia/hj6gjlGvvCDZeQ2ehLrPqxP+zU1g/K1O+N76c3RcjqL+X17t9/OJZdqoThk9xgx96yWNIcw+d8grvbxwrnB2PBhSqYNKW78LyJv/d+5heV9z7mpyuoN4OdoJTQja/YeRwG17yAKV/w1/npkBhePw6Di/L5mLa210Rs5oThi0fjg4vzMEi4n9/8K2HclH6oq1qtIUoBn/WTlfKlOL75K0D+PCX38RXs+CsXowd3ArfBm+sX9y9r4eTfx+QvrnNy1WdmY1z7jZe7LwZOxxXLDnAvhSXBco6e+KprElZ9zY/hr9aFo/3c2eheyL9T4Mb8eKdY/CKM+QGsbAVWGIKv3e0N8lx00Tcm1dvD0ysXOyePwPjt14XJ2tuhY2p1mIClk1vi3r6fMWr4YAweOpy9hxDUHfAbZvVTbZap/dF4/NhFpTM9p25DVodZGKn2n2ca9pqI7xzDsXgEn2byumR8NmUQa6xVfVKcLtIHg/41kwruxbAclDUruOZdErgX1Z7nlkX+f2tSHDrl415eyy66jtyyuGn+IoR83KYDQ4RdnZL+ayZ94TyyVyaFbzDhzrMtUmi7yV9QZKeD+tz7oyu/Y/aDHvJ/68OXU3T/aG37fBR3P4SK/0rGClI6ukCBxr9dKvFYLN26KNBV3+SYmBao53+lY4qjJDqzqHrpMrY5DBm7pfKvmTQxZStsuFBwu/eKu3Ft6JSPFZ7i6qhVYQr5/itBKk04QStKQLjzRbUb95yrNO6dL6fo/inOWHEUdz/E20Dp6AKtlHgsvp666KpvtNXzbdExJdGZRdVLl7HNUVpjt5QMFkEQBEG8XkppSZAoCW9uuYZ4XyAZI94VXsOSIEEQBEG8XshgEQRBEEYBGSyCIAjCKCCDRRAEQRgFZLAIgiAIo0C5SzA+IQFlyvAfZiNeL69evaK2Jl4rJGPEu4K6LCsNVnp6BiwsFP/jjXidUFsTrxuSMeJdQV2WaUmQIAiCMArIYBEEQRBGARksgiAIwiggg0UQBEEYBWSwCIIgCKOADBZBEARhFJDBIgiCIIwCvQxWWlwoHnDfzSbeCbivj8pyhYCO8F8rJt4lGFk6YkNCEBqVwvdtrgzPs+WnSg19dQcjTUFU9BOSufccvQzWi5h92OqXJISKhkEUdnj0wZFI+fvJhIG8ij+C8e7u2HnzpRBTclJ9l2PQ0gt4yv7Oyw7CmoGDMOVADH9SB17dO4VZgwZhg+8TIYY3YCU1esTbQ15mENZOGIFl2/fg4JV7rGGQwX+LJ7zGeOMuO3S5cbzTYwn8Hhk2jkuiO17d2Y/Pe3sjhrt+2VT4TDuE6Je6XV9dxol3B70MVg1bF9wNjiJheAPEBvsg09wcJ/1v4bkQZwhiM1cM/n0Rvu/eQIgpnjJ1O2PGL8sxoBX/UTVOme0dMhinomlSYqw8Dr0EX6tvsGjFIswf2QrlYYo2/X7FTwt7o55ISFQK6Ks7xKbWsO92GdHxQgTxXqKXwRLXtUdL/wgkZGlXUIxMCik74y7t5YT3Hc4wXD+Sh/7ffY2Wp/0R9lR3A8F5QFyfaFtSqVjdFlYSIaAgVyZPL5XK5EFu2fA5/5NFgirWdWBuxv3m0mWB8/eys1TpCeOBlw0ZROVE8t+KcSuSVINt9Qp8oDDyyUlxFKc7FHJa0Fu3gF3jOrgaGSeEC1KUjBPvBvoZLFNbOPf8F7ej8wupDPHn1mB0v3GYuXgxfpg4Gn/5JsuVGWE4L0Kv43ylzmjaoilafHwRV4LThTOFk5cZii1zR2DApPlYvHgGJs7eg7Bn6soiHZfWDdXwjh4EbsWskf0xef5iLJr5DRbsjcDVI17Y4M8vAXKGc9+QPxD4iEFeRih2L9uDK3l5OOe9BEuXnpEvIRHGQjpueC/FhqPheBF1GOuWLsZvZ2LlZx5dXYuvD4bLf2uDk5PvR47mx/rk/hg1ezvCi5lEFaY7mOxEnPpjLL4YNx2Ll/6I8eNW4eoDzRlvjYaueHk9Fo+EsILiZZx4Z+D++S3Hkyfpwi/dSLm0jPlmV5gQ4pFG7WMmei1nriTnCTFZTIj3TKZ3797M4QhFHFHStuZ5yVzf/iWz8NRDeSj92npmwA//MKnyUGE8Yf5d7sHM3RXOSIWYV6mXmCXD3BjPJeeZDHkMn0bRP6+SzjLz3KYxh24/k4c5Ei+sZkZ6uTG/X0yTh/OYSGZHn8XMlVQ+Dxf+Xx936uO3CH3Gs0omeLg4r538GM/f5zlxh5lpWsZ6/jK0UVB3vGTzTmbGrrjApL7iY3Kz2OtN78P07bWbiRYuwdVh53BVHXh0kXHCmFGXZb08LA5LdrZjGZaoNtuRIcJ/F6oNHor21RWL3hI49fRAV3EpLoK/p+RJQ+G/vx46utSQhys1bYePgi8h/GHhM8nc5Bs4f6k3Bn3ugPJCnIlVJ/T1KPx5VVKID+I9hqCXk2opqFbXXvjMUm9RId45ZLh1aRusvLw0x3rfIejmH4Q7T4r2bvLrjjxZFPx3lcPgQV1gZcLHiSX26PX5J3xAQAQb2HcLRHyiyjvTR8YJ40VvLSSuwbr20iBEK3cNSZGeyqBRHV6hKhBVtEB1ERksQ8kMDYCPvTNqVXouX6d/kWMN+x4R+MefX77RBpOZjri2trD6QLP9Lapq9pE66elhcK1fA2WFMIcI5qhSi/qQUMCP9ZjTqzFn3hzlMW/BdgQwPniYKiQrhPy6g3mWgUeMPWrX1JSxClWqC78UmMKucXf4RapkXh8ZJ4wXvQ0WN9txaBuNO3FSIYZHJqOH7qVPOq77nkV2xBZMGeQJT0/uGIYVJ3IQ4xNS5DMj5kXJ+0MqK3qGTBAczTsPwMAvBqodXpi0cBFaFjO50a47cljdUbzcVbR1hHVApIbM6yPjhHFi0DpPHdu2OBcWI+zKkcCihgVux92XhxTkpaUinv9GJKEn3LKHr29X/LDjKI4dO6Y8jhxYjB6ZR3A1RPu2FlElCziERCIh34PwlIeRwq+CVK/VBiH+IRoPtvOyExEblieECIIf62kmFnBxcVE7HODg4ILalYRkRaCuO0QVzVGzSigSEvlzCp6kFNwRKLa0RgPxDSQk8TKtj4wTxotBBoub7ThejBJmO6Zo2rEPZNvW43BYlvw8I0vFxZ17kFBZHiT0JCnMH9EfusKpsubMVWzqjFY9THAxOELrVl6TGm3RvVcwNm3ywSNhm/DT0P3Yf77wbrdq8wXcX/yJlevP4E6qFGmJQdjzxx5EVy5q0mGBKk1FiLt/X+ftzYQxw4/1jD2bcSKSH+vcMmH8mTUYN20PYnTYFqyuOzg57vC5Gbas24toYbv7i+RL2Lk3GvlVB++dPcTNaH6HrD4yThgvBvWq2NIezWwCWNeeF7IyNn0wZZ4TApYPhpubG/qOXonMT2dgkLPwJJUoMQzi4Xf8Orq2bqZ8qKyOY+s+KLvfH+EvtBkUCdoO/wE9c3dhpIebvE/mnKuJryZ1F84XRGRqg8Hz16C7eRS2rV6MlTuDUXfQV+huXXgfilANXYd8ifT94zF04N+4Vcg7NsS7AzfW533fHDd/58e6m5snFl6sjJHTPdDgAyFREeTXHY3cp2Bq49v4eZC7vLzxfyShx0wvNNSyYcvOoTMCWe+Mf3G+5DJOGC8ibqsg90P9u/klIeroVJyqMB2TPlR/yMm9TJiDsmYSmJKtKoC+bW0I8hd/c8tCIjEVYkpCOs7+MhK3W/+NSV34/25RGNwjTFN9LkGUKv+ljHEv7L4ykaC8/EVybpLF/RunGdiTW3AZuUKHKfhrVje556RNd3D/dOB5TtFymicNwp+Dw/DZ4WFooGbPDJNx4m1FXZYNNli5D4JwIs4KPTrV1dhZRhTOmzBYupL7yBe//nwVHedOVW5Zzrx3Cn+MP482fyzHJ/ULzniJt483LWOcEdO2TA1WSygMiv66Q4qwMz4QteyBxlVIHt91StVgESXn7W5rGe4cX4Pf917Ei0rOsJSFIja7Cfp/OxFD29BWYWOBxjPxrkAG6w1jFG3N/Y+4bH6ObCqRkPdsZNB4Jt4V1GWZttIQ2jExhYQ1VNxBxoogiLcBMlgEQRCEUUAGiyAIgjAKyGARBEEQRgEZLIIgCMIoUO4STEpKhpmZ8OYf8VrJzs6mtiZeKyRjxLuCuiyTh0UQBEEYBfQe1huA2pp43ZCMEe8K9B4WQRAEYXSQwSIIgiCMAjJYBEEQhFFABosgCIIwCshgEQRBEEYBGSyCIAjCKCCDRRAEQRgFehmstLhQPMgUAgRhlMgglcqE3+83r2s8M7npiA25j+dCWB+eJUUh4bH+/VQadSDeHvQyWC9i9mGrX5IQIv5LXsUfwXh3d+y8+VKIUfHqzn4MWnoBT4WwLqT6LsewXeFC6P0h7vgPGDJwFQKfyN+bf695bePZJAdRB9Yj8KH+bWyS6o95R8ML+dy+DpRCHYi3B70MVg1bF9wNjiqRYiRKh9hgH2Sam+Ok/y2aNRqATbfJWPbrMDSrIhJi3l9e13gWoRrsXJJwMzpdiCk5ZvUc4OIThbt62pvSqAPx9qCXwRLXtUdL/wgkZPFSxMikBZZXmGztcc+zhQBRYhhE4fqRPPT/7mu0PO2PsKdFj2KZVApZrvwX2xcF+yM/XHouHZ+nIIrzz/MXw31OXyi7sDScjBToe7V8HMr6yuM10/PypOXaAsrzWuSLK5eboTPZ6bgflyL/LTKzgLW1RYGvKRdVzrtK/vHMtpi8DTS8GqFPCsYV0iECdRp0QmBYjHJypU0HcP2jNY4tWmRuC6f6AbgTp3/dDK0D8fagn8EytYVzz39xO5rvzdwHZzFn3CHEKPWnFAHbvDDi++NqM6N0nFs9FDtuZAlhoqS8CL2O85U6o2mLpmjx8UVcCS5q1piOS+uG4vjNCOyaMRYzFy/GD5P7Y9Ts7QjPZ+jyRKnw3zgT4+cvxsKZ32DAl7NxJFK15Jj7KAgb547A2JmLsXjpj5jQbzR+OxGrVBCvYo9i1OozOL9/Hr5l03BlDB6tWcajq2vx9UHNpUcu3+jVV4SZPV/f/T5++H3SZCyaPwmeX4zCet97iDq6AGOnqco9cUdVLoNUBPw9D8PHTufr9sUw/LxD5X1yRn7fkD/w76V9mDpgOL6beEEuk5rXVpUz7Ivv8JOWct5l8o9nBg9wYu63OBmpkpOMG1sxyPN7/Burikv1W4kxG68V2UZmDZzR/nQoEl7y+Z6G7sSolaplawbxODx3CMZsVpUj77Nv5+DifYb3kFzFrIeULJwred0MrQPx9qCXwQIksGvcFn6RsfKQiY0L2lv6K2dBebI4hJyqiPIZ15GQJMRJ4xB5uTWaNJTIw0RJkSEi+Bxse7ZDPZEETdv2gP/5QDwSzhbGsQ1nUH/qFqxZtAi/btiNSU3DseDPixrLP892HUdK+3nYsGIRlqzZiAW98rBr90Wh7HRc9t6AtObzsHHNIixa8As2bByEp+v3IPCRajA/v7IX4VW+ladZsmYb1gw3x451J0q8lHP0WhaGrlyLxSs24n8/dcLl5axyKjdGKJetW/ds7DqrMiR3jv6GLRkf4dct6+R12+g9B3Yh87HB54mQglM+ATgSaIXvDx3FweOeaKBlFZArZ3NGJyzdtwnL1crZ4fc+TLA0x7MINmjauQquRsbJw5zsxYb/A3OLDKXh4CalseEBaN28McoLMdoQS2xh3/EyouP5cCUHF7T2i1R6c3nJcbj5pBLKnlIZlNz4SATmtUMjW76j7BzaKZcs9albadSBeDvQ02ABlg1dYRmWKFdqnBA5ts3E5VBeiPISohDU9kuM65+tXDvOjgmFb7MWaGhFAqAPedJQ+O+vh44uNeThSk3b4aPgSwgv5mFyzZ690b66os0lcOrpgVb+ZxGslk/S1QPdnCoIIVM07vIJGtzIAO+IWaDLhPWY1c9OuXwmrmGNuuKreJQmRLCUNe8Jt2583ThqNmuHpgmxSE0vun756d6tPaxM+N8V6tnBXtQObZpV5yPYujm2bI+8TH75J08WCp+/y2HwoC7KPGKJPfoO8cC16+FKoyxCffTxUKXJD1fOxb9fwMPjE9QWPiHFlTP4550Y2kLRLu826uOZw86pPVLO35JPODgPJOpCK4wa/znSBcPBTUpDTzZF84YW8vSKpVT1g/fALeDQtA5uJ/DGRFzRHs6dzuBmmFQeTosOwtPuE/BlL5VBSWbjZB/ZsxMzPiyuzy1ZxuKBYExKWrfSqAPxdqC3wRLXsIWzNAjRwizbrvFHeHQ9Vi7wsZH+aNzOBS0d2iNSWDuOjbwEFzbOSp6aKCmZoQHwsXdGrUrP5crgRY417HtE4B92IBeFUz1r4RePqKItHJqFIUNtG3OZmlVQWfgtp6wpJOoDNVeKuzfO4sCW1Zgzbw6mjZ2Ow7l5wkkeU6fqMC8wuHOQU8jzsMKwqKy7B848y0AK7uPEn3Pl9VIcC/93BdkXWOWrtJU1YF5J+KkFrpzHjGuB2bTITILy78k3EPOPZ85IdMwMkq+QcN7GDdbLcWntiBaRvGfCTUoDnduhcS2+zW4fGQZPT0/lMcRzldIDr8EaQ4UxkRuPFq44x+oFmdwTisBnzVvC0clR8JrSEXkrFB1cHFUTJPmSpcqYlLRuHIbWgXg70NtgcV6VQ9to3InjZylm9i5oFRqEO49SEBtUTj674QSrxelQxL+MQ2RALaV3QJSUdFz3PYvsiC2YMkihFIZhxYkcxPiEFLnsliPTeBTNIkVWCd65ka/v//gtlp1Jg5VLF4yZMBs/r/kFA030Fp1SRcw448P+AzHwC9XhOWwCflrcAWr6SgdyIJMV0ZDvOPnHs9jUHi7dIuQrJJy3YdHMnlXzNrDvdkb+rIublNbv6KycgDoPPIBjx44pj8PHZqO9sJpiYuOgNCYcNZ3aoe7FKCRkc55QRzS04T28MuyEN0Uah6jLvdG8sak8LQ+3ZKkwJiWvG4fhdSDeBgzSOnVs28pnKZxK5GdBgQjz88OtjBaoX1OknBmFX4zArTw+jig5uck34OvbFT/sOKqhFI4cWIwemUdwNaTgO1kKohIV6/o8eWmJuBPbGlaWQkQxcOnDbnXC+MkD0NnVBfWrsV4HO44Lv2Lh5NxLUS7TcTxJUTyH0A9RRXPUqpIKkyrOcHFxUR7ODg5wcKlT5LMVdcSW1VC/SigSEoUIgWeJobid+P5sElIfz9zyq13j7rgadhH+wZnsBJSbbPJxlyN9EBmUK8QVjwjWsGnHGxMOzptraumPsNMRuPGZM+p/wOoKwcO7dTkEfkKcOlWs7fFSWMHRp26lUQfizWOQwapo6whHdpbCz/AlcGzRHcc2bUZiWwdh7ZefGW1f8xdyO7vQerCeJIX5I/pDVzhV1mxAsakzWvUwwcXgCEHJFOTx6R04HMYrXUaWigvb2f7p/jGa6fgsUWRWHpVFkbgTw8+8ueXBsEN7CywJFgencKr6ncA/15PlS5oPQ/dj12nFExP94O6//ee52LPlEKKFnY+MNBH/rP0K83YX3ib5EcEeLT83w5Z1e5U7KPMygrB7xQqEp70/i0Ka4xmo1NgV7U9vhXdoW+VyKZemlvca7EjvDCedNyRo2dTRzQSbN/0PDk4N5BMLhYe3es1BdGlWcCMH5yG5JqqWLEteN8PrQLx5DDJYYkt7NLNRvSPBCUxrsQjtHezkYQ7Lxu3YOAu0dVLFEbrDLcn5Hb+Orq2baR1Ajq37oOx+f4S/ELRMPnqO+xoVzk2Fm5sb+vSbhhNlBmHeyNY6D0axxBV95zohYPlgeRnuHhPhW7UzPEq4JFjGpgcmjLXC6Z/HyJc0F52rBI/PWwtn9cfefR4mOUdjiZc7f4+es3G1wnBM6Fuy5w9cOfM7JmCFUI6H1zpkffgD+jb9QEjx7pN/PHO76xw7mqB8N8UElDUclk3Qjo2r1k4VpwvccptFQKTSGHI7/yzgLH90oMCxeU9UQTutO4k5Y2LfLVC1ZKlH3QytA/HmETEs3A/17+aXhKijU3GqwnRM+pCeT+mKvm1dMtJx9peRkLodRB8HkfwlyFcmhmwi4F7EzEFZMwlMC9ltpxsyyHJNDSxDG6VUP+6F0+zSuM83y9s2nrmJ1/7J3qg9axZUu1ZLRkbgBkyM7IjtXo2FmJJRGnUg/nvUZdngJ+cNXAfDtqxM5+UX4s1gKjF0x5spJGwZhivx12GsOEqpfialdZ/Gyesaz5yH1GqwI8q84J8h6UMl+y7wsszReA5aEkqjDsSbxWAPiyg5b8LDIt4vaDwT7wql6mERbysW+Hj6ITJWBEG8M5DBIgiCIIwCMlgEQRCEUUAGiyAIgjAKyGARBEEQRgEZLIIgCMIoUG5rT0pKhpnZe/Kvqd8w2dnZ1NbEa4VkjHhXUJdleg/rDUBtTbxuSMaIdwV6D4sgCIIwOshgEQRBEEYBGSyCIAjCKCCDRRAEQRgFZLAIgiAIo4AMFkEQBGEUkMEiCIIgjAK9DFZaXCgeZAoBwsjhvtRLH7QjtMPkpiM25D6eC+HSgpGmICr6CX34lSgRehmsFzH7sNUvSQgRr5M8WSjW9/kOJ2Pl73cryU0+h1nu7th6NUuI4WEQhc0efPpXd/Zj0NILRX6hNe74DxgycBUCnwjlc5+IJwNGKDDJQdSB9Qh8qCl/xcN9QLQvjkRqz8eUTYXPtEOIflnScon3Gb0MVg1bF9wNjtL7U9WE7ohNbeHS8z7uJKYLMTxp0UEIYxicC4vRmKXmxkcipFJ7NLLV7cONNt0mY9mvw9CsCp/+VexReA08iBjSIwSLCNVg55KEm9Ga8mcoYlNr2He7jOh4IYIgdEAvgyWua4+W/hFIyNKu1WRSKTtLl+J5thBBGIAEdWyaIJA1TKplGSliwwPQqlsXWF6Mwl21bkhmDdmjtjaolc9eMdlCn+RznkRmFrC2tkBZ9jcjY9NkvWS9tJfI0tp/nPfFl6PrUk5hssBdSxGnSCPL5cP5UdS9KM9PeZ3Cksg9x6LLIFT9q94XdRp0yid/aijbtSiZ0Nb2FrBrXAdXI+OEsCbF93nJZZEwfkx+ZOF+cP9gsFw53f5ZpsikAmSZKxBVri9capYRYjkhi8fR32Zizi5/RAX748iO3XhQuTWa2lSAiZCGKFlbc5QXZ+Ly6gdoPKgpOEeIQTwurXmIJuO/QKWLh/Hc5TM0lHtIUoSdXQ+m8Qh0ql8OeWnhOBhRFg0zduGndRcQdPkA1q07gsw6bdCyXgV52a9iDuHLbY/xSUcbPLuxCyu8fZGSloCk2Ju4kV0fne2ryNO9iD+FZbNmYK/fHVz/Zzf+OnQP1Zs3R/3Kqv5XpzhZeOS3EhOumaJh+GrM/vsKgrm67QqAqH4bNKlpypeBVFz9eynmrDmCkNvX8e+BNTgSVhFNWjWCuXBZxXVmbzmHqJCz2Ln6EFKtmqGpXWWlzD0I3Iolc1bhxK0Q+B5cg73+OWjY0gVWZrp5ocZISWVMvX9v+B/Dzu1X8LxOc7jUKgeTirlIWR4Ci8+bw6qMqs3U2/Xa2QPYczFDrV2zEed3BNm1m+LmylnY4nerQNtzcv3PiWdo8aEdOy3jyd/nZw/sQMBTW7RyqSGfVHEo6rrj/B3cPFu8LBLGjYYsc//8luPJk3Thl26kXFrGfLMrTAjx3D+7gPFccp5JFcKvUi8xK2f/yVxLzRNiCI6StnUeE8nsGP4dcyKGb8ecuMPMd17eTELeS+b61r7M7xfT5PG5L0OY9e6zmYsPhHRR+xi33l7M6nNJ8jBH4qXfmBFDtzFh2ao0XJ9lyEN8uG+v3Uy0WpflPr3OrB42jTl0+5kQw5Zz4Vdm5ARvJjpbiMhHcbLAyY+71yjmL7W6ZYTvY6a6LWKuKNJcXs18/cMBJuG5PMi2Qwpz4icP5f0yTBbj99cIZu7WYEYqxHDXWTpsokZbTfNazlxJVtxQFhPiPVPjnt9FSiJjeS8jmb3TpzObr6j6Qhq8hZUTtn/lzfaE+Xf5COZwhKINeVn7m5XJPbcVAvCSCd33M/PzoXAhzOXxYEaMWVlo23NyvXP4YmV/c0hDdjBfTtjGRD/n43JfRDK7FsxhjgvXVsii9438ssjKiaoY4h1CXZb13tZu2dAVlmGJeCSEOaTSdIgtJMrZkolVJ3y36Gu0snp3Z7L/BSLYwKHbPdyMTpaHuWW/tK72qCUyhV3j7srlmrx7UbhazxX1a6rau6x5T7h1qyGEgJptPka7zAfIKMEuz4cBRxD+4RD0cuK9Mo5aXYein8UphKmvR6qhiyyYsnXrrla3yo494PZJIC4E8fdZrcN4/PnT56hXTh6UP0+p01CMGw/487nJATh2ohncPVxQXh7DX2fyusXoWoe7jgy3Lm2DlZcX2ldXXFcCp75D0M0/CHcUG03ec0Sm9vhi+XJ82V7VFx/UsoPd01ikPubayAIOTevgdgLf7nJyZHieLoGFhPeG2d5Ek/7zMM/DUQjz1OzZW7Pte3+BDv5nEfyQkcu1fbdAxCeqlv1k2VJkVaqC8qZ8HrGZPQbNXYheDnyYk0W/5n3g1lwli7W7jscfSz4rsAxOvHvobbDENWzhLA1C9CPVoG/46Vj0ur8Og8dMw7I//4fzQffwvJDnEkRJ4A1TWFgsMpCO2PBQfOTUQL5EUtHWEY3/CUXCSwbJcSEwa+eAemoD19SpOszVwiJTU0hKOLDT08PwPHAffpw3B3OEY+68VTgW/wTxSdofxusiC2bNbPMpGQns7J0R9URV5rPEIJw+uAUr2WvOnjYGs3eplBuTmY64Ni6oX1nzhkwlElbhcb+kSE9lEHN6tbLe3DFvwXYEMD54mCpPTrAwsnSE+57E9j+Xytto/NRfcF3NntdgJ6jpahutxBJX9J1RE94ThmPCnFXYduwi7jx+KZxV4VTPWvjFI5LUgW2zMGHCxMu1X2Ss/BxH5VYe+Lb6MYz78hv8sOpvnLoUgQy1x1icLLZ2aaCcoPCwMs32uWLJkHh30dtgyWf9baNxJ04qxPCzocELN+B/88egQ0MRIo4vwPAZexDzgmayhsIZJiefKCRmJiLqQms0suV9F7GlPZo3v83OUtMReSsU7R3s5PGljXWrTzDwi4Eax7hpC+Gh5nWpo4ss5GXLCjwwl7IzbAVRx3/AhJ9PI6uqMz7+ciLm/LQWi4cqZvQ8zIvCHsqraN55QL66e2HSwkVoSVNyOXmZQVg7YRL2BOehYQcPjJsyG7//Oh0t1ZrHxMYBLSIjNTZa1WY94E37VuC7z5ujSloQfp8yBmt81LwwlhxZgR5Glpp3z8m1dUCkcuMQ50V3mbAee1ZMZ72omnhwbQ0mjf4Ffimq60plpE/eV/Q2WBx1bNtqbqvOlUGWa4qK1vbo+PFQfDN/PoZU2I3LISolROgHb5h8cfNkMAKdXdFQWFqTbzt2FeN6qA/iLn+KJg01FXppYFG1OeKzy8PRxQUuaodDI9a7qVrI9XSQBVlkItQcdJZ03I+KQ4va3NIU+zsiFJ+Mm4p+nVuiiW11dhbNXiuPT8khsqwOlxBWiT5VL0SGtGjW43zMGTIJLGpYIM3EQqPeLi4OcHBwQe1KfI73nbzkePhaDcb4Cb3Rjm0ba4uC3ooI1rBpdwa3o9WX72QQmVVDA9fOcBsxFSsmdYLPAT+NXatRiZoGLC8tEXdiW8PKkg+LLa3RQHwDCUmKTKzcsJcwq2aDVp17YvS03/FVZ38c9+d3E1av1QZhrEemvmOR8w5jQ+I0PDHi3cQgg8XNjhyV26ql8N/8NeZtC0SGsPSTnRyF2zcbonZ1xZMMQl94w2SJvTv3wqqlHayEeA47h3aI+XszLrRzQP0KhnkNokoWsBfHITExS7ntvFarHmjntw87/JKVk5NHN3Zi9oDF8NO0OAK6yUJO9kns9o4QlI8MDy5uwY7QT9G1mQUbNoWkkgiRkXFK5ZR55wj27lbN2E0sW+DDXsHYtMkHjwRllXnvHP6cegB3ZZzKNUXTjn2QsWczTkQqXrCWIv7MGtY7ZL29gitY7ydly6LGvRBEC14MI0vFlQMHNZYE5cu1jdsql+9e3TvF9v8KnEkQ2jVXyvZVBKo61YH6I+vHp3fgcBifhiv3wvbNSOz+MZopJ1zcSs1D5Xte3Ivsw347jgfCvCYvMwqRtyuikXV1ebhq04/QPngjNp9XyCLbn+c2YPaBaOSU/lyNeMswyGBxs/5mNgG4E8dJtgRtveah/eON8PJwg5ubGwbNOIf6s6bik/qGKVGChzNM5rBAGwdbIYZHXJ/1YkzEaNLMAZWFOH0xqdEJQ77MxfZvB2HM5mtyYyGu5IpRi93w6uQsfM72K9e3Y/+8i65LJqK91g01uslCpa4T4FXzHMa7c2n6Ye4+E4xYNBLN5M+kJGj9+RzUDVuGgcI1Z5+siq6e6luX2esMX4AB5ucwpR+fZthcX9Sf8zW6CMt9ZWz6YN73zXHz98Hy825unlh4sTJGTvdAgw/kSd57ytj0wLcD8rBpjLu8jfqOXonUxh9rLAlycButLITluzJ1e2DyvFo4P19oVw9P/H3vQ8wa2krj+VLPcV4QnZgiT+PebxQOiQZh3sjWGmnsHDorNw7Z9J6CSVZXMcuT70+PISuR/uEPGNCSX3qWy+KiUfjgvEIWPbHItyq+/+ZTjUkc8W4i4rYKcj/Uv5tfEqKOTsWpCtMx6UPVDiPupT/uATv38JsehBZE37b+T8mVIcfEtED/8X1bll+e04HCZCHVdzmm3e+N7YMb8y8R5xReJvdS8CsTCcoX8VoRX0bRMqdLOe8KeskY9xJwdk6hbci9/7d/sjdqz5oF1c4/7gXeHIiKadei2j5PGoQ/B4fhs8PD0EAoVtGfZc0kMC3kJU6uzBy2prrKImGcqMuyQR4WRwPXwbAtq/nwXMQKGe3aMXK0GCsOvm91VxC6yILItOgy5bv+ijEyfBlFX0eXct5r2D4vqg255btWgx1RRmOjC5+nuHYtqu3FEgd0/bYqZOmqNUhFfxZmrDi4MslYvV8YbLBMarvCvVNdMk4E8R5Qv5UHWtcv7bVUCZw+7YnGwv+zJIjCMHhJkCg51NbE64ZkjHhXKNUlQYIgCIL4LyCDRRAEQRgFZLAIgiAIo4AMFkEQBGEUkMEiCIIgjALlLsGkpGSYmdFLKv8F3AfJqK2J1wnJGPGuoC7LtK39DUBtTbxuSMaIdwXa1k4QBEEYHWSwCIIgCKOADBZBEARhFJDBIgiCIIwCMlgEQRCEUUAGiyAIgjAKyGARBEEQRoFeBistLhQPMoUAQRgA/9XYksB94Vb9A4KEoRgynpncdMSG3Jd/3l4fniVFIeEx9SehG3oZrBcx+7DVL0kIqXh1Zz/c3NyUx4jxczBv2V/453qy3gJNaPIq/gjGu7tj582XQowxkY6zv/TFSp8n8tCre6cwa9AgbPDlw7oQd/wHDBm4CoFPhK/Tcp91JwNmEAaNZ5McRB1Yj8CHqq8F8/B9rcjr7jYKM+bNwdrdFxGt6DsWk1R/zDsaXmDSkpvmiwV93JX5PcZMw9x5q3DAJwIZuUIi4r1DL4NVw9YFd4Oj8FQIq1OG6Y8Fu73h7e2NNT+MQr/ONXB941hM/vUiHglpCP2JDfZBprk5TvrfMvpJQJm6nTHjl+UY0KqKEFM8Nt0mY9mvw9BM+Drtq9ij8Bp4EDH59SWhM4aMZxGqwc4lCTej0/kM+fjw27XyvLu9V2D80M9h9/wcFgyfjiOR/ITLrJ4DXHyicFdL/4mYtpi6erc8/47l32GIhxOe/LsMY2bsQcwL6vD3Eb0MlriuPVr6RyAhq6DQiPABKkgkkLBHhWq2aNbOA7NX/YLOKSux5Z9kIRWhDwyicP1IHvp/9zVanvZH2NPiB61iyY2RSVlPhDsK80Y4T4VPU2CJTs2L4crj0jzX1amR5y3suhJUsa4D83z/8k5R1+fZfJi7pkyYVYvMLGBtbYGy7G95uqyXbLu8RJZaeqJkGDqe6zTohMCwGK0TKFG5CvK8EokF6tq7ovuoBVgyozYOL9qBcNboiMxt4VQ/AHfitMuyqeLaFnXg6PoxRi9cjq+s/8Fvu2+VcCmZeBfQz2CZ2sK557+4Ha2b1hKb2aPvEA8EnvTTOpMidONF6HWcr9QZTVs0RYuPL+JKsPZZrQLOwO0b8gf+vbQPs8dPx+LFM/C152gs2BuhMdhfxJ/Cwq/6Y/L8xVg48xsMGLMYZ2JVS46cFzNq9Rmc3z8P387k0wwePVs5Sy6Mp7d2YtZIvtxFbJ6f9t6CVK3/FfULfKSIlOHO8WUY3W8cZi5ejB8mjsZfvhG4tG4oTkXzabi6jF59Re4NPL61D8t2XUSu6BJ2LVuM387EytMQJcPQ8WzWwBntT4ci4aVug7tWpyH4wvYYroVIeQ/NVcx6aLpNZrn03QYNhflhf4RrMbDEu41eBoubGds1bgu/SN0VxAfWdmiaEIkHj0nI9EOGiOBzsO3ZDvVEEjRt2wP+5wOLXWZlEIAjgVaY8uc6LFq0Fls3D4dox1/4N5bvh9w0f/zx43k0mbgLG1YswpI127B1XC2c/GkzgtU8uOdX9iK8yrfYuIZPs2a4OXasO1HoBIR7BrF6XjDazuDLXbxmE0ZXu4H9vq+EFAV5HroPi/eKMWoTW/6iRVixYRU6PNyF7aHaH1pYtRqBn8d+hjLMJxj78yLMdbcTzhAlw7DxLJbYwr7jZUTHCyeLgTM6dRqb4HzsPXnYzqFdoUuS2hDXsEZj8T9ISBQiiPcGPQ0WYNnQFZZhiTo/lxJbVoONmH/uQJScPGko/PfXQ0eXGvJwpabt8FHwJYQXeNidH3P0dOsCKxM+ZGLVAm073UP8Q947Swo6hVvN++BTpwryMEflFl+gb6szuKjmwZU17wm3bvy1OWo2a8cqrFikpmu/flLQv4jyGIJeauXW6toXbg2EihSANcg3DqDh4KFoX10hJxI49eyFprpqMkJvDBvPFnBoWge3E3Rf8q9eq43wiy2rPrckGYsHOnponMGr0UFv1UUYMXr3uriGLZylQYhWLucUDZORjsQ83dISBckMDYCPvTNqVXouf77zIsca9j0i8A870ItCBAdYWQgBNXLy+L/p6WFo7dIA5fmggAQ1alXBjQcqBWTqVB3mBeYbOcgpZMcWV65r/RryZ00KRKzxrFKrsEmLFOmpDBrVURlFDlFFC1QX0UTndWPoeK7BGrz0EnhJ6alRwi/22vIlSd09NAYZSAkTBJh4r9DbYIlgA4e20bgTJxViiib7fiwiK9uhdlVSPiUnHdd9zyI7YgumDPKEpyd3DMOKEzmI8Qkx+LmgVPZ6JhL6lCuT6bqbgyhNDB3PJjYOaBEZqXXjRn4YpCI27AnaKCcn3JKkI65GxgnhoslLjsOdJx1Rq5oQQbw3GORX17Fti3NhMcXu1mFyE3Hm0D7U7dcS9chelZjc5Bvw9e2KH3YcxbFjx5THkQOL0SPzCK6G6P9OFrc0ExYZq7HDS6FQPrSrK8SUHIuqzbWWW/jMmPWk6lrgyk3NDSF5yYmIIs/8P8GQ8SyCNWzandFp40bmjZPYG/MpOrioXP8q1vZ4eT1WhyVJKa79sxMp3bvASXi1gXh/MMhgVbR1hONFzXcoFFuM+a3M6bgbchab5s/FKckkjPegh+L6kBTmj+gPXeFUWXOAik2d0aqHCS4Gayr5klC1VQ90Cd6IzeeThTKkSDizAweeeqJbM9Xzp5JSiy23vVCunFwpbu7ZiwumhRufxh+PQT2f37FibyAS06VIunMW67edZmtUOKJKFrAXxyExMYu2tRuIYeNZ+8YN5kWWkFeKrNQ4+B5ahinzb6Pv7JFopibPnIfmmlhwSVLxGgV3JMX5Y9/yyfgt9DNM+7JVvmVs4n3AIIMltrRHMxvNdyheifZjnrBsNcjzO6zYE4Yqn8zD71O7wkpIQ+gOg3j4Hb+Orq2baR2gjq37oOx+f/k7LfogNnPG6F++xgfnZ+Fz+X8V8MSyK5UxYcYA1PtASKQH4kquGDXfExl7xvL/6cBjIgJrjcF3H6k/1dLExLIdpq+aBqcn5/Dnb4vx58lMfDxuJFoUsVnHpEYnDPkyF9u/HYQxm68Z/cvUbxJDxzO3ccMiIFLD4J1f+62whO2JoXPWwT/ZCRM2LUIfB03h4pYk7bsFaixJMqIA/DphkDL//HUX8NRhDNYtGogG5ci7eh8RMSzcD/Xv5peEqKNTcarCdEz6UPNhOVE4+rb164abzb4ykaB8vhd5DUVeblm2XFMhogRw72rt9JiBiksPs0quCCWVK0OOianGJo/3mTcxnrnJ1f7J3qg9axZUOz11JyNwAyZGdsR2r8ZCDEFoyrJBHhZHA9fBsC3LKgshTBgv3H8VKG1jxSEvVwdj9TyCnc3P2YNotQf3jwP9cb1Sd9S3FiIKg4xVqWDIeOa8pFaDHVHmhX4bZyrZd4GXZY7OOw2J9w+DPSyi5FBba4fblHF163psOHgNZRq5wCQ1FMkftMW4WVPxqZ0B65PvISRjxLuCuiyTwXoDUFsXDfc/Ap/Lp/hlIZHosY5IkIwR7wzqsmzwkiBBlDYiU/4fnpKxIghCHTJYBEEQhFFABosgCIIwCshgEQRBEEYBGSyCIAjCKFDuEkxKSoaZ2Wt4CYcoQHZ2NrU18VohGSPeFdRlmba1vwGorYnXDckY8a5A29oJgiAIo4MMFkEQBGEUkMEiCIIgjAIyWARBEIRRQAaLIAiCMArIYBEEQRBGARksgiAIwijQy2ClxYXiQaYQIAjCILgvMr/JD6CW1nhmctMRG3Ifz4WwITDSFERFPymVdnmWFIWEx/p9VDI/pPveLHoZrBcx+7DVL0kIAbnJ5zDb/U/cUvtSLMfzW1vh7jYRJ2M143Mf+WKRlvSEdnLTfLHQ/TvsDXspxKiTjrO/9MVKnydC+M3Df9Z+Cfwe6d6/r+7sx6ClF967r82+uncKswYNwgbfN9d/pTaeX8gQdWA9Ah+qziee+QFD1l3LZ8RkCNr2Ob6YeBB384lI4sXl8vTSsqnwmXYI0S8VCfg8sw/HCmEe7qOfJ+d7aLkGELq7nzy9Sao/5h0NVxq/PFko1vf5ruB9sPc9y90dW69mCTE8nDxv9uDTk+57s+hlsGrYuuBucJRSuYhr2KKZ3VUkJAoRArGRl9jOjsfN6GQhhudZQgRut3dA/QoiIYYoDkYUD++lOxD+ggT9XaJM3c6Y8ctyDGhVRYj57ymt8WxToTrsXJLY8+nCGbbshq6wCoiHmg2TlxF5IRey+OtISFKXZynuRwWglVMDVDC1hn23y4iOF07BFHaNuyM+8p7GpCYvLQpBwXl4fjIUCUrjxl8j4lpFtHGwhVk9B7j4RCmNo9jUFi497+NOoqqeHGnRQQhjGJwLi9Hw7HLjIxFSqT0a2YpI971h9DJY4rr2aOkfgQRhliBCbdg1y8TVyDh5mEMuMAF56NKtPW4HR6oJmQyx4f/AxqEuKgsxRPGIGGe0bBKKP7YE6rzkwi01SdnjuZbVEMUyFPd1Xy6NVFryJRMmW7e88i8IZ/O/FXWS5fJhbSjK1VZvjqKuW+x95cq05NOMK1hftfTy/NqvLaeI81xZ3H1z5SfHKZbOJKhiXQfm+f/tX3HXKUVKczzXadAJgazCV8iouLYtHJ/64U6cyphwBuBaXkd06BimYdzyZHEIOVkHjawt2JAFa6DqaNShYh072PhFKuvJ8SwuAreadULbBpfUjBtbVnIcgmPboL41ez/mtnCqH6BWBwnq2DTRqCdnLGPDWWPZrQssL6qMG0cya8getbVBLdbGkO57s+hnsNgZinPPf3E7WjGY+NlPXNg9PBJi8tISER3bCZ8N6oAmakLG4AFigyvJZz5ESZCg8+Cv0TlhHbYUs3zEZMfjyPKxGDBpPhYvnoGv+o3Gb0djlbNGbolj35A/8O+lfZg9fro8zdeeo7Fgb4TGzLIwuGWYgL/nYfhYNu/Sxfhhcn98s/Q47mpbsWR5dHUtvjoUiNA9szB25mIsmj8JA76cjT038y+95CDy+AKMncan8ew3Cut9VTPU4q6ry329ij2K0auvaMzSc9OuYtWgQ4gRlBRX37H7L8Lnj/GYsvhHTBjUD+NXXsHd6COY/80kVbkHVG3K8SBwK74fORozF/N1GzV7O8KfKjRfOi6tG4r9Pj5YNmYQvv7ufwh+xCjrHKi2fMqVM2tkf0yer62c0qc0x7NZA2e0P63ydpTezN0UeZjjSWIUnnT9BEPad9AwGnkP4hBRmfdkODjv7OX1WGUdeG/mNFuWEMFyPy4AzdsNRM92JhpG49n9WMQLnowI1WDnKtbwdriyq56KVHp+DBIR5++ITl/0Riuxv5pxY72++Ntyr688GyLd92bRy2BxytOucVv4RarWkyvaOsL5KisAQudwM5+wLvaoU9MezZurXGb5zCemExra8GFCd8Rm9vAY3RsJG/+GX4piQOWHVYx//YzrdSZj18YVWLRoLbbtHIcK53/GTrW1eQYBOBJohSl/rpOn2bp5OEQ7/sK/+dbctfHoyl5su9ccC9azeRcswooNm9Bb9jcOBRRuSDNPrYef1SRsXbMIi1dsxLbvWyBw/h8az7meX9mBwPJjsFFI8+eMJgjYcAbhgvLT5bqG3Jc6WXv8gIGrsWbRL9i0+0c4By/HzAPlMH4dX+765Z/h8Va2boK8v4o/gpWrH8Ft+TY2zyL8umE3JjUNx4I/L2oYx0MnktFv7REcPjYb7a0KLgspyum+cB82rFCVs3DLFZ0965JTeuNZLLGFfUf1pTy+7MDIe0L9eU+mbcO6cqNRX225kPNk0rrao57QLCY2DnBNDEK0ICMi2MCxbRXW8PCGSeHJNLGtAzuHzmrLhbwn06SZg9KTsXNop7GUx5Xd2vKG0jBxXl9gngtsrW3h0O2ecrmQ8/pCTzqieUPO6+Mg3fcm0dNgAZassFmGJapmP5Zc55wWZh68wDg52cEc1dnZjaVy9pMWG4K7PZxR/wNaw9WH8o3cMdo9Ddt2+CjbXp3c5Bs4fb4Z3Hs6yGeEHOJKrhjg2RqnLwSqKU9z9HTrAisTPmRi1QJtO91D/EN+oKqW1FSHYhmvWofx+POnz1GvHB/mZrB1Gopx44Hmer06puY90b1bDSEEVHbsAbdPAnEhSJWnLJvGTS1NzTYfo13mA2QIu7J0u27R96Urkh4fo1V1XkY5JWznJEbbdi2V5ZZzbInWJhmQvuBCMty6tA1WXl5oL+ThFJtT3yHo5h+EO08EjczSa0AvNCz0+QVfTpk+fdG1zgdCnATO/X/GxjGtlf35Oii98WwBh6Z1cDtB1Sdc2fVO8V6XwgA0sbOQe0xNLRXeTDp7jVB8xHoyZflscgNl3y0Q8YkKb4YzPJ0Rez1GXk+lAajHe3Yd/XlvhvNkIi/UZY2MSpbE9bmlvFg8ECY/XNmcYVJ4XQpjWUvEe0wKzy/vXhSu1mMNa01Vn5Hue3PobbA4YXOWqs9+OLfbDv7R9wsIDDe7eSR37WW4H+OrdK8JfTBFo/5f4aOUldh5vqCBYDLTEdfGBfUraw6KClWrw+xyChQOjQgOsFJMGtXIyeP/cstinp6eGsepaJXifcbOfE8f3IKV8+Zg9rQxmL1LpVS0YdbMVv4MQAU7U7V3RtQTlSExdaoOc7U0IlNTSPKN7eKuW9x96YqJeYUSyKgU6akMYk6vxhy2Xopj3oLtCGB88DBVSMZiUVki/NIGX04LBzshLGDCtoPEVAi8HkpzPHOeU3r+jQkNfBF9V9MA8B5Tptxo5MkSEXWhNRrZqrcPbzzUvRnOMHUKDUECOwlQNwC8Z3dGbjQ4Qxae1065tMjBL+UV3MQRFhaLjHzGkvOYGv/DG9jkuBCYtXNQen0cpPveHHobLPkMpW00OzuSCjF857wMiUNKUqyGwHCzm1ahsUh8Gsd2prp7TegD1/b9pkxA+sq/cPFhQUPBvCjaeOhCtU4zcOzYMY2jjwPfn1HHf8CEn08jq6ozPv5yIub8tBaLhxatUPOyZQWej0mzVbKjC/pc97+keecBGPjFQLXDC5MWLkJLTUtdLDKZ4f1XUkpzPHPLbS0iVc9ueMMkxu245AIGgPOYwuPuIyshCoHOrmiYb5mUMx7WAZGqHX6sYXLo6Ic7d6X5DADn2Tkj7G6ifLffPbWlRR5uKc9R4zkXV7aTTxQSMzWNJe8x3WY9u3RE3gpF+3wTCNJ9bw69DRZHHdu2GltA+c4Jgo/PdQ2BUTyovHUxAhGVNd1rQj9ManyEMVPKY/Wv23Hjmcp9EFlWh0sIqyzyPaRPvhsJWc/8Xo4+pON+RCg+GTcV/Tq3RBPb6vzsvxgPRhaZqPTueNhyouLQorZq2aZo9LuuNmRhKk+Tg0lLQXSeZnuVDAksalggzcQCLi4uaocDHBxcULuSkKxYLFC9rgWr2O8LYZ48VqGGltILuUVRWuNZBGvYtOO9HQWcYQoJvoRLwSEaBkCxSeNsxC1YtbSDlRCvQGxpjQbiG2rb33nD5Bt2AiH5DAC/SeMWLsX6aywtKqhiba+5iUNumHxx82SwhrHkPSYxrof6IO7yp2jSsOCkiHTfm8Egg8XNUBzVtoDynROIHTuv5BMY/kHlgU2bkd5W070m9Kd21xEYZx0O35sqrW1i2Qbd+93Cpk0+eCToi8x7Z7Fz+1MM+LhpKSxHmEJSSYRIdqaqUKCZd45g7+78/pMmOdknsds7Qsgjw4OLW7Aj9FN0babrjFO/6+ZHXNsGXTNP4dTpWGRIpUiLv4Sth6+hssgQoTRF0459kLFnM05EKja2SBF/Zg3GTduDmEJ2T2rDsXUfyLatx+FbfDncf4+4dvA3/HkzrYACLm1KbzwX3JjAP2PaiX2XPtMwAPxS3nFs3XRX6+453pt5iPzvdpXZuxMnK2kaALlnl7gNOw60UnpL6hTcxMEZJkvs3bm3gLHkPKaYvzfjAusNantninTfm8Egg8XNUJrZaL7fwHWOCTiXWFNguAeVTqxSyO9eE/rDDbhuQ0eiQxX1UWCKFkOXYoj5OUzp5wY3NzcMX3AZtb+Zij6NFA/yDUGC1p/PQd2wZRjIls2VP/tkVXT1LCOc106lrhPgVfMcxrtzefph7j4TjFg0Es3yPWsrHP2umx+xxBWfz+2GB7snw8vTEzM2JqB1396oZqAiKWPTB/O+b46bvw+W183NzRMLL1bGyOkeaFCCZufK+WFZR9xex5fj7jEM25M/xMwBzV67wSrN8cydt8i3lOfY0QRmHfMbAN5jMmG6FLp7jvPO1Le/c4anHSvz1fI9W5Ibt24m+KBpwaVFDu48t4kj/1KeOVuH/MaS85g6mog1dhqqQ7rvDcEIPHmSLvwqGZFHpjC/n0sSQoQu6NvWJSXvZRaTlfVSCJU+L7OyGOkLIVAEKZeWMV47w+S/S6NOul63OF6+pqYptfoZUM6bHs95TByzd9Ji5kpynhCjP7lZ15k17tuYaMOLYtKvrWe8tvOyaCik+/4b1GXZIA+Lo4HrYNiWLfhAnXjziEwlr3V3malEgvL5/0NDMZRGnfS5rjZMX1PTlFr9SqmcklBa45nzZloNdkSZUtgAJJY4oOu3VSFLV3gz+lPJvgu8LHM03o3TF9J9/z0izmpxP9LTM2BhYS6PJF4v71tbp/oux7T7vbF9cGMhhnjd0Hgm3hXUZZkM1huA2pp43ZCMEe8K6rJs8JIgQRAEQfwXkMEiCIIgjAIyWARBEIRRQAaLIAiCMArIYBEEQRBGgXKXYFJSMszM/uOXPt5TsrOzqa2J1wrJGPGuoC7L5GERBEEQRgG9h/UGoLYmXjckY8S7Ar2HRRAEQRgdZLAIgiAIo4AMFkEQBGEUkMEiCIIgjAIyWARBEIRRQAaLIAiCMArIYBEEQRBGgV4GKy0uFA8yhQBBlAa5MjzPFn4T/ymGjOc3kZfJTUdsyH08F8LE+4NeButFzD5s9UsSQprkZUZh/x8zMNLdDW5ubvhy/C846HePPiNtALlpvljo/h32hr0UYtRJx9lf+mKlzxMh/Pbx6s5+DFp6oYjPksvgv8UTXmO8cVfxFXTWgEmlhn9enSgeQ8ZzUXmfRhzB79PHyPO5uw3Dd8v34Np9lQzrndckB1EH1iPwoUJYNCnuuoTxopfBqmHrgrvBUQUUUF5mENZNXIzoGkOwdLc3vL29segrGwSvn4cNvm+vQjUGGFE8vJfuQPgL7YPUuDFFm36/4qeFvVFPxMe8ij0Kr4EHEfMu3u5bhiHjubC8GYEbMHHJTdQfukCeb7f3Lxhuewfr5vyBwEd8p+qbV4RqsHNJws3odCGHCl2uSxgvehkscV17tPSPQEKWpgA8Dr2Eyw1HYPSApqgmkUDCHrWc+2Pqd93wODK2iBk2URwixhktm4Tijy2BOi+FyKRS1kuR4rkWR4U7x82SGRmfRndvhvN8+LxK5N6QtriCZTLZ2uskklSDbfUK8t/yOmW9BIOXyOLS5l8qFK5XeJ0V56WQ5QpRAor7Lr6MotuPR3UdjXtXo/gy3jyGjGftedNx68opNB/xNfo2rSHPJ5FUR4v+U/D1h5mIiuMNjSF56zTohMCwmHxjQbe8hPFi8iML94P7j7jlyun2351FJhUgy1yBqHJ94VKzjBALSO8H4ERERXzarREqCjNlDrOazdDVtTbof0fzlKStOZgX93D5SBK6zPKC+Znf4W/WGa3rlRPOZiPO7whSa/dB2/p8HJMdj6O/zcScXf6IunEKO1cfQkIFJ7S0rwIT7jyisGfAVjyslYwtSzfg8o2z2Ll2P8LKOKC9k5U8TWEwuIfDM2cgvn4fOFTlOzkj6G8MnbAbVdp0R8MqfFzqlV/x7XExerStC5O0cByMKIuGGbvw07oLCLp8AOvWHUFmnTZoWY83Uo/8VmLiVQv0dbbC4xu7sMLbFylpCUiKvYkb2fXRma07x4PArVgyZxVO3AqB78E12Oufg4YtXWBlxl/3RfwpLJs1A3v97uCG/zHs3huDaq7NUadiGeV9Z7rkwXvGEpwKvYnjO9fg1F1LNGtmh0qCKOc+CsLfS2Zj3fFbuBV0Dvv+2IvYymz7NeLbj0NxnR3n7+Dm2d3469A9VG/eHPUr84Vo64NUq2Zoale5yPYtLf6r8aw9bzbuBhxBTKVP0blRZSGOwxTWzT6EizUvp4bkNamYi5TlIbD4vDmsyigqp1tewrjQkGXun99yPHmSLvzSjZRLy5hvdoUJIZ7cF5HMjul9mHFLvBnfiCRG+lI4QWhQ0rZ+9fgSs8BtEXMlNY+RRu1jpnktZ64k5wlnnzD/Lvdgfr+Ypgxf/H0EM3dXOCMVYnKfXmfWfzeM2RLwTB7OYyKZ//VxZ8auuMCkvpJHMa9Sz7LXmMiciFGUWziRBycz3x+KEUIvmetb+zJeXl7MwlMPhbgsxu+vvso65bB1duvtxaw+lyQPcyRe+o0ZMXQbE5bNX4+TJ6+dKnni8vTttZuJVqtOTtzhfPeexYR4z2Q8l5xnMthQHpPCnPiJbQuN66xm5q+9LJzn7nsYM/J7b+bOM76MvFcpzL+/jWBmsu0lk8ew7ffHWGbJ/hghzLZN0lnmB6H9Obj2XD1sGuN9g29PjsQLvzIjJxxgEuRJuPvX7INXqZeYpcMmMocjsoWY18t/OZ615eXkdIrbKOaXXReYqJQsZVvmR/+8nNyPYNtTTUBYdL0uYTyoy7Le29otG7rCMiwRj4Qwh9jMHoOXbMGYJhn4d+1MePYbhok//4nTt+j5VWlRvpE7RrunYdsOH422V5CbfAOnzzeDe08HlBfixJVcMcCzNU5fCFRbljVHT7cusBKm+yZWLdC20z3EP+SXTVRLhapDsbxm59QeKedvyTdIMIhH1IVWGDX+c6QLzyPyZHEIPdkUzRta8BlYypr3hFu3GkIIqNnmY7TLfIAMnXeJyXDr0jZYeXmhfXXFjFoCp75D0M0/CHeecMtKUmSlMzCX8F4bR+1O4/HjNx2gmm9noFOfXmhYgS9DZFINXT088GKXH6JfcmVYoMuE9ZjVzw5l5SnY9qthjbriq3iUxocfsrN4v+Z94NZc7Tpdx+OPJZ+hFltsbnIAjl/phkGfq/rAxKoThgypjH9D7gsxbxeGjGdtecs36o9ftn+HRk8v4fdpg9B/2DdYuO44wuX9pEL/vBZwaFoHtxOShTCPrtcljBO9DZa4hi2cpUGIzvcgU2Rigea9v8JPq7dhn/cv+KqDGS7+NglrfDQFi9AXUzTq/xU+SlmJnecLtimTmY64Ni6oX1ltDYelQtXqMLucAkV3ieAAK5U9UZKTx/99dHUtPD09NY5T0XxmcX17dMwMQkISg9z4SNxwaAeX1o5oERkpfx6RlxCFQOd2aMxpbwFTp+owV6uSyNQUEs0qFoMU6akMYk6vxpx5c5THvAXbEcD44GEqd0826D1uKCLXDcbYaYuxYecZ3LyfJeTnEaEJbOpIhBCPuLYtnMSpKuOZK8XdG2dxYAt/rWljp+NwrtAwLOnpYWjt0kBpjHjY+5FI5EaO64PbCMDOhXM16rr26F2kxd1/K5/lGjKeC8srNneG+9c/4M/t+7FzxXdo+cFVLJ+0An4pqnSG5K3BGjvFJEkdXfISxoneBotTDg5to3EnTirEFOQDSXU4ffQlFsxxR9AmH9WWZcIguLbvN2UC0lf+hYsPCz7NZ14Y/oS/WqcZOHbsmMbRx4G3MGJTe7h0i5Dv0kqODoJFM3t2vmsD+25ncDtahthIf9Tv6AwreerSpXnnARj4xUC1wwuTFi5CS8E4cjPsJRt3Y+6YbqgnisehhYMxb3eE2qaILEhfCD8V5MhY/42H8xgP//gtlp1Jg5VLF4yZMBs/r/kFA000h4pUVrQwm9Zoi+4a9RyIQWOnYk6/pvkM3duBIeO5+LymqFDNHt1HLcDUXhHY5RMrxBuW18TGQTlJ0k7heQnjRG+DxVHHti3OhcUIykCGG7vm4fd/7slDBUhnCt1JRZQckxofYcyU8lj963bceKaa/Yssq8MlhB3ETzUHcfLdSMh62sqXrAzHFHaNu+Nq2EX4B2eieUNuqY+Puxzpg8igXCGuNJHAooYF0tgZv4uLi9rhAAcHF9SuxKdismWs9yZBXft2+GzwV5j/wzjk7r6IcEGpMUhAYoqmcsx7EI+gSraoVlWEvLREhN3qhPGTB6CzK+upVpOgvCmg/hZP9VptEBYZq7FDjZFxL7PGIYO1fKJKFmgcKUWlhs4adXVu5IAmDasolxrfNgwZz+p587JDsfWHJbiYoN2QZL7SnFDpm1cEa9i04ydJHCW9LmF8GGSwKto6wvFilDDTMoVtwzoI3rUZh249UQrzswf++N/G/6HOsJbKd2yI0qF21xEYZx0O35sqg2Vi2Qbd+93CJnYG/EgYn5n3zmLn9qcY8HHpze4rNXZF+9Nb4R3aFo1s+Y7l5KGW9xrsSO8MJyFOXzilby+OQ2JilrCt3RRNO/ZBxp7NOBGpWOaTIv7MGoybtgcxrEXJkwZh7dfTseOq4hmLDEmRoYhpaA0r5fqjOa7v+lu5PJSXGYo9m3agbr+WaMAmEZmVR2VRJO7ECEYtV4qwQ3s1lgSrNv0I7YM3YvP5ZEHO2Xqc24DZB6KRwxo3E8676nUNe7f7KfsgLyMIW+d98Va/4G3IeFbPKzazhX3tO9ixfS/CHgmmnm3HhICd2LrTEv1a2vFxAvrnlbCTpLbwYycPHCW9LmF8GGSwxJb2aGYTwLrz/OA3bzUCP46pj5vrRuBzN/7N+CFfb8KTtksxa0CDt3ZmaaxwL1B2GzoSHYSt5DymaDF0KYaYn8OUfnwfDF9wGbW/mYo+jT4Q0hiOWGILx44mKN/NQam4TCyboB0bV62dKk5fTGp0wpAvc7H920EYs/ma3JspY9MH875vjpu/D5bfl5ubJxZerIyR0z3QgL01scQVo+Z3R8LmEcL5fph/zhLTZ/RWepbcs7thU7ohavWX8jQeQ1biXvOfMNODV2ZcGX3nOiFgOX8Nd4+J8K3aGR5qS4LcJpZRi0bhg/OzBDn3xCLfqvj+m0+FZVAJ2g5fgA/L/KPsAw+vdfJxMLYLvz3/bcSQ8ayZl7//wXXv4NdR/eX53Dw8Mfnvu+i2ZH4BOTQkL7dpwyIgUjCyJctLGB8ibqsg90P9u/klIeroVJyqMB2TPtRcAlK8oGkqPIgmVOjb1iWF2+n3PKcsJBJ22m+s5MqQY2JaQIY4+XplIkH5Ql414uVP896597B2eRyEzcZZaG8l4ssoyy/5FYR7KZgtwUwC0yJenNJ2HQ24F5Szc/7zcfAmxrPWvML9o6g2YtE3L/fMcf9kb9Sexfapcvcoi47XJd5+1GXZIA+Lo4HrYNiWZZWKEFbACbZi1xTxZuCe5Rj9YNVirDg4+SrMWHHw8lf0vcvLKDQJv+uvKGPFUex12Pob0zgwZDxrzSvcf3F9oW9ebtNGq8GOKJN/o5GO1yWMC4M9LKLkUFu/GfJ7WO8yJGPEu4K6LJPBegNQWxOvG5Ix4l1BXZYNXhIkCIIgiP8CMlgEQRCEUUAGiyAIgjAKyGARBEEQRgEZLIIgCMIoUO4STEpKhpmZbh98IwyD+yAZtTXxOiEZI94VVLIM/B/XPr0FjSjenAAAAABJRU5ErkJggg=="},"148df854-4102-44ad-aa3d-44cee56dc924.png":{"image/png":"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"}}},{"cell_type":"code","source":"# Note: Name format 'Braund, Mr. Owen Harris'\ndef get_title(name):\n    title_search = re.search(', ([A-Za-z]+)\\.', name)\n    if title_search:\n        return title_search.group(1)\n    else:\n        return 'None'\n\ndef feature_construction(df):\n    # Name Length\n    df['name_length'] = df['Name'].apply(len)\n    # Title\n    df['name_title'] = df['Name'].apply(get_title)\n    df['name_title'] = df['name_title'].replace(['Lady', 'Mlle', 'Ms'], 'Miss')\n    df['name_title'] = df['name_title'].replace(['Mme', 'Dona'], 'Mrs')\n    df['name_title'] = df['name_title'].replace(['Don', 'Sir'], 'Mr')\n    df['Age'] = np.where((df['name_title']=='Master')&(df['Age'].isna()), 1, df['Age']) # Special Treatment\n    df['name_title'] = df['name_title'].replace(['Dr','Capt', 'Col', 'Major', 'Jonkheer', 'Rev', 'Master'], 'Special')\n    # Cabin\n    df['has_cabin'] = np.where(df[\"Cabin\"].isna(), 0, 1)\n    df['cabin_category'] = np.where(df[\"Cabin\"].isna(), 'nan', df[\"Cabin\"].str[:1])\n    # Family\n    df['family'] = df['SibSp'] + df['Parch'] + 1\n    df['is_alone'] = np.where(df.family==1, 1, 0)\n    return df\n\n# Feature Construction\ndf_train = feature_construction(df_train)\ndf_test = feature_construction(df_test)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.897916Z","iopub.execute_input":"2022-07-24T10:47:47.898176Z","iopub.status.idle":"2022-07-24T10:47:47.936928Z","shell.execute_reply.started":"2022-07-24T10:47:47.898144Z","shell.execute_reply":"2022-07-24T10:47:47.936049Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Impute Data\n\n# Impute Fare by group mean\ntemp_cols = ['PassengerId', 'Fare', 'Pclass']\ntemp = pd.concat([df_train[temp_cols], df_test[temp_cols]], ignore_index=True)\ntemp['Fare'] = temp.groupby('Pclass')['Fare'].transform(lambda x: x.fillna(x.median()))\ndf_train = pd.merge(df_train.drop(columns=['Fare']), temp[['PassengerId', 'Fare']], how='left', on=['PassengerId'])\ndf_test = pd.merge(df_test.drop(columns=['Fare']), temp[['PassengerId', 'Fare']], how='left', on=['PassengerId'])\n\n# Impute Embarked\ndf_train['Embarked'] = df_train['Embarked'].fillna('S')\ndf_test['Embarked'] = df_test['Embarked'].fillna('S')\n\n# Impute Age by group mean\ntemp_cols = ['PassengerId', 'Age', 'Sex', 'name_title']\ntemp = pd.concat([df_train[temp_cols], df_test[temp_cols]], ignore_index=True)\ntemp['Age'] = temp.groupby(['Sex', 'name_title'])['Age'].transform(lambda x: x.fillna(x.median()))\ndf_train = pd.merge(df_train.drop(columns=['Age']), temp[['PassengerId', 'Age']], how='left', on=['PassengerId'])\ndf_test = pd.merge(df_test.drop(columns=['Age']), temp[['PassengerId', 'Age']], how='left', on=['PassengerId'])","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.937943Z","iopub.execute_input":"2022-07-24T10:47:47.938170Z","iopub.status.idle":"2022-07-24T10:47:47.990474Z","shell.execute_reply.started":"2022-07-24T10:47:47.938142Z","shell.execute_reply":"2022-07-24T10:47:47.989755Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.Age.quantile([0, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1])","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:56:09.908503Z","iopub.execute_input":"2022-07-24T10:56:09.908805Z","iopub.status.idle":"2022-07-24T10:56:09.921425Z","shell.execute_reply.started":"2022-07-24T10:56:09.908772Z","shell.execute_reply":"2022-07-24T10:56:09.920429Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Mapping\ndef mapping(df):\n    # Age\n    bins_1 = [0, 5, 16, 28, 36, 50, 64, 100]\n    df['age_cat'] = pd.cut(df['Age'], bins = bins_1, labels=[0, 1, 2, 3, 4, 5, 6]).astype(str)\n    # Fare\n    bins_2 = [0, 8.1, 15.1, 30.1, 50.1, 100, 600]\n    df['fare_cat'] = pd.cut(df['Fare'], bins = bins_2, labels=[0, 1, 2, 3, 4, 5]).astype(str)\n    # Sex\n    df['Sex'] = np.where(df.Sex == 'female', 1, 0)\n    return df\n\n# Round Age and Fare\ndf_train['Age'] = round(df_train['Age'])\ndf_test['Age'] = round(df_test['Age'])\n\ndf_train['Fare'] = round(df_train['Fare'])\ndf_test['Fare'] = round(df_test['Fare'])\n    \n# Mapping\ndf_train = mapping(df_train)\ndf_test = mapping(df_test)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:47.992134Z","iopub.execute_input":"2022-07-24T10:47:47.992384Z","iopub.status.idle":"2022-07-24T10:47:48.016114Z","shell.execute_reply.started":"2022-07-24T10:47:47.992351Z","shell.execute_reply":"2022-07-24T10:47:48.015149Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plot_kde_survivors(df_train, 'Age', 'Survivor Age Plot')\nplot_kde_survivors(df_train, 'Sex', 'Survivor Sex Plot')\nplot_kde_survivors(df_train, 'Fare', 'Survivor Fare Plot')","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:48.017474Z","iopub.execute_input":"2022-07-24T10:47:48.018161Z","iopub.status.idle":"2022-07-24T10:47:48.671176Z","shell.execute_reply.started":"2022-07-24T10:47:48.018112Z","shell.execute_reply":"2022-07-24T10:47:48.670512Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Drop Columns\ndropped_cols = ['PassengerId', 'Name', 'Cabin', 'Ticket', 'SibSp', 'Parch']\ndf_train = df_train.drop(columns=dropped_cols)\ndf_test = df_test.drop(columns=dropped_cols)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:48.672401Z","iopub.execute_input":"2022-07-24T10:47:48.673258Z","iopub.status.idle":"2022-07-24T10:47:48.683202Z","shell.execute_reply.started":"2022-07-24T10:47:48.673210Z","shell.execute_reply":"2022-07-24T10:47:48.682347Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Catboost Prediction\n\nSee [this notebook](https://www.kaggle.com/code/odaymourad/learn-overfitting-and-underfitting-79-4-score) for parameter tuning.","metadata":{}},{"cell_type":"code","source":"y_label = 'Survived'\ncat_cols = ['Pclass', 'Sex', 'Embarked', 'name_title', 'has_cabin', 'cabin_category', 'age_cat', 'fare_cat', 'is_alone']\nnum_cols = ['family', 'Age', 'Fare', 'name_length']\nfeatures = cat_cols + num_cols\n\nX_train = df_train[cat_cols+num_cols]\ny_train = df_train[y_label]\nX_test = df_test[cat_cols+num_cols]","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:48.684830Z","iopub.execute_input":"2022-07-24T10:47:48.685179Z","iopub.status.idle":"2022-07-24T10:47:48.697299Z","shell.execute_reply.started":"2022-07-24T10:47:48.685136Z","shell.execute_reply":"2022-07-24T10:47:48.696385Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_test.info()","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:48.698785Z","iopub.execute_input":"2022-07-24T10:47:48.699108Z","iopub.status.idle":"2022-07-24T10:47:48.721348Z","shell.execute_reply.started":"2022-07-24T10:47:48.699066Z","shell.execute_reply":"2022-07-24T10:47:48.720225Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def run_catboost(train_X, train_y, test_X):\n    cv = KFold(n_splits=5, shuffle=True)\n\n    model = CatBoostClassifier(random_state = 42,\n                               # iterations = 500,\n                               # n_estimators = 10,\n                               verbose = 10,\n                               cat_features =  range(len(cat_cols)),\n                               early_stopping_rounds = 10,\n                               learning_rate = 0.01,\n                               grow_policy = 'Lossguide',\n                               auto_class_weights = 'Balanced')\n\n    params = {'depth': [4],\n              'num_leaves' : [8],\n              'n_estimators': [6],\n              'loss_function': ['Logloss']} #'CrossEntropy'\n    \n    grid = GridSearchCV(estimator=model, param_grid=params, scoring='f1_micro', cv=cv)\n    grid.fit(train_X, train_y)\n    \n    print(\" Results from Grid Search \" )\n    print(\"\\n The best estimator across ALL searched params:\\n\", grid.best_estimator_)\n    print(\"\\n The best score of the best_estimator\", grid.best_estimator_.best_score_)\n    print(\"\\n The best score across ALL searched params:\\n\", grid.best_score_)\n    print(\"\\n The best parameters across ALL searched params:\\n\", grid.best_params_)\n    \n    pred_train_y = grid.predict(train_X)\n    pred_test_y = grid.predict(test_X)\n    return pred_train_y, pred_test_y, grid.best_estimator_","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:48.724405Z","iopub.execute_input":"2022-07-24T10:47:48.725106Z","iopub.status.idle":"2022-07-24T10:47:48.736747Z","shell.execute_reply.started":"2022-07-24T10:47:48.725059Z","shell.execute_reply":"2022-07-24T10:47:48.735755Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Training CatBoost\npred_train_catboost, pred_test_catboost, model_catboost = run_catboost(X_train, y_train, X_test)\nprint(\"CatBoost Training Completed...\")","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:48.739229Z","iopub.execute_input":"2022-07-24T10:47:48.739765Z","iopub.status.idle":"2022-07-24T10:47:49.017380Z","shell.execute_reply.started":"2022-07-24T10:47:48.739726Z","shell.execute_reply":"2022-07-24T10:47:49.015657Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Feature Importance","metadata":{}},{"cell_type":"code","source":"# define predictor and response variables\ntrain_pool = Pool(X_train, y_train, feature_names = features, cat_features =  range(len(cat_cols)))\n\nmodel_test = CatBoostClassifier(random_state = 42,\n                                # iterations = 500,\n                                verbose = 10,\n                                depth = 4,\n                                num_leaves = 8,\n                                n_estimators = 4,\n                                loss_function = 'Logloss',\n                                cat_features =  range(len(cat_cols)),\n                                early_stopping_rounds = 50,\n                                learning_rate = 0.01,\n                                grow_policy = 'Lossguide',\n                                #auto_class_weights = 'Balanced'\n                               )\n\nsummary = model_test.select_features(train_pool,\n                                     features_for_select=f'0-{len(features)-1}',\n                                     num_features_to_select=10,\n                                     algorithm=EFeaturesSelectionAlgorithm.RecursiveByShapValues,\n                                     shap_calc_type=EShapCalcType.Regular,\n                                     train_final_model=True,\n                                     # plot=True\n                                    )\nsummary['selected_features_names']","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:48:19.672609Z","iopub.execute_input":"2022-07-24T10:48:19.672925Z","iopub.status.idle":"2022-07-24T10:48:19.812298Z","shell.execute_reply.started":"2022-07-24T10:48:19.672892Z","shell.execute_reply":"2022-07-24T10:48:19.811363Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Evaluation and Submission\n\n\nThe evaluation method refers [this notebook](https://www.kaggle.com/code/kevrchen/catboost-gridsearchcv-feature-engineering/notebook).","metadata":{}},{"cell_type":"code","source":"import scikitplot as skplt\n\nskplt.metrics.plot_confusion_matrix(y_train, pred_train_catboost, normalize=True)\nprint(classification_report(y_train, pred_train_catboost))","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:49.157279Z","iopub.execute_input":"2022-07-24T10:47:49.158087Z","iopub.status.idle":"2022-07-24T10:47:49.359153Z","shell.execute_reply.started":"2022-07-24T10:47:49.158052Z","shell.execute_reply":"2022-07-24T10:47:49.358448Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission = pd.DataFrame({'PassengerId': PassengerId,\n                           'Survived': pred_test_catboost})\n\nsubmission.to_csv(\"submission.csv\", index=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-24T10:47:49.360446Z","iopub.execute_input":"2022-07-24T10:47:49.361344Z","iopub.status.idle":"2022-07-24T10:47:49.368375Z","shell.execute_reply.started":"2022-07-24T10:47:49.361297Z","shell.execute_reply":"2022-07-24T10:47:49.367765Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}