{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nimport gc\n\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.preprocessing import StandardScaler, OrdinalEncoder\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import accuracy_score\n\nimport lightgbm as lgb\n\n\nimport warnings\nwarnings.filterwarnings(\"ignore\")\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n for filename in filenames:\n print(os.path.join(dirname, filename))","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-11-16T10:47:21.467224Z","iopub.execute_input":"2022-11-16T10:47:21.46769Z","iopub.status.idle":"2022-11-16T10:47:23.366028Z","shell.execute_reply.started":"2022-11-16T10:47:21.4676Z","shell.execute_reply":"2022-11-16T10:47:23.364415Z"},"trusted":true},"execution_count":1,"outputs":[{"output_type":"display_data","data":{"text/plain":"","text/html":"\n"},"metadata":{}},{"name":"stdout","text":"/kaggle/input/amexfeather/test_data_f32.ftr\n/kaggle/input/amexfeather/train_data.ftr\n/kaggle/input/amexfeather/train_data_f32.ftr\n/kaggle/input/amexfeather/test_data.ftr\n/kaggle/input/amex-default-prediction/sample_submission.csv\n/kaggle/input/amex-default-prediction/train_data.csv\n/kaggle/input/amex-default-prediction/test_data.csv\n/kaggle/input/amex-default-prediction/train_labels.csv\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# Load Training DataSet","metadata":{}},{"cell_type":"code","source":"train_dataset_ = pd.read_feather('../input/amexfeather/train_data.ftr')\ntrain_dataset = train_dataset_.groupby('customer_ID').tail(1).set_index('customer_ID', drop=True).sort_index()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:51:49.77723Z","iopub.execute_input":"2022-11-16T10:51:49.777722Z","iopub.status.idle":"2022-11-16T10:51:56.956219Z","shell.execute_reply.started":"2022-11-16T10:51:49.777674Z","shell.execute_reply":"2022-11-16T10:51:56.955122Z"},"trusted":true},"execution_count":3,"outputs":[]},{"cell_type":"markdown","source":"The dataset of this competition has a huge size. If you're reading raw CSV files, It will create a out of memory error. That's why we read the data from AMEX-Feather-Dataset.","metadata":{}},{"cell_type":"code","source":"print(f\"Total space by dataset = {train_dataset_.memory_usage(deep=True).sum()/1e9:2f}GB\")","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:51:59.450388Z","iopub.execute_input":"2022-11-16T10:51:59.451146Z","iopub.status.idle":"2022-11-16T10:52:00.003877Z","shell.execute_reply.started":"2022-11-16T10:51:59.451093Z","shell.execute_reply":"2022-11-16T10:52:00.002405Z"},"trusted":true},"execution_count":4,"outputs":[{"name":"stdout","text":"Total space by dataset = 2.776791GB\n","output_type":"stream"}]},{"cell_type":"code","source":"del train_dataset_\ngc.collect()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:52:02.232303Z","iopub.execute_input":"2022-11-16T10:52:02.232771Z","iopub.status.idle":"2022-11-16T10:52:02.400543Z","shell.execute_reply.started":"2022-11-16T10:52:02.232702Z","shell.execute_reply":"2022-11-16T10:52:02.398979Z"},"trusted":true},"execution_count":5,"outputs":[{"execution_count":5,"output_type":"execute_result","data":{"text/plain":"53"},"metadata":{}}]},{"cell_type":"code","source":"train_dataset.head()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:52:04.796828Z","iopub.execute_input":"2022-11-16T10:52:04.798016Z","iopub.status.idle":"2022-11-16T10:52:04.832684Z","shell.execute_reply.started":"2022-11-16T10:52:04.79797Z","shell.execute_reply":"2022-11-16T10:52:04.831446Z"},"trusted":true},"execution_count":6,"outputs":[{"execution_count":6,"output_type":"execute_result","data":{"text/plain":" S_2 P_2 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 2018-03-13 0.934570 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 2018-03-25 0.880371 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 2018-03-12 0.880859 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 2018-03-29 0.621582 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 2018-03-30 0.872070 \n\n D_39 B_1 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.009117 0.009384 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.178101 0.034698 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.009705 0.004284 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.001082 0.012566 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.005573 0.007679 \n\n B_2 R_1 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 1.007812 0.006104 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 1.003906 0.006912 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.812500 0.006451 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 1.005859 0.007828 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.815918 0.001247 \n\n S_3 D_41 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.135010 0.001604 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.165527 0.005550 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... NaN 0.003796 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.287842 0.004532 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... NaN 0.000231 \n\n B_3 D_42 ... \\\ncustomer_ID ... \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.007175 NaN ... \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.005070 NaN ... \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.007195 NaN ... \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.009941 NaN ... \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.005527 NaN ... \n\n D_137 D_138 D_139 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... NaN NaN 0.007187 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... NaN NaN 0.002981 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... NaN NaN 0.007381 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... NaN NaN 0.002705 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... NaN NaN 0.002974 \n\n D_140 D_141 D_142 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.004234 0.005085 NaN \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.007481 0.007874 NaN \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.006622 0.000965 NaN \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.006184 0.001899 NaN \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.004162 0.005764 NaN \n\n D_143 D_144 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.005810 0.002970 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.003284 0.003170 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.002201 0.000834 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.008186 0.005558 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.008156 0.006943 \n\n D_145 target \ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.008530 0 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.008514 0 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.003445 0 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.002983 0 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.000905 0 \n\n[5 rows x 190 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
S_2P_2D_39B_1B_2R_1S_3D_41B_3D_42...D_137D_138D_139D_140D_141D_142D_143D_144D_145target
customer_ID
0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fbac11a8ed792feb62a2018-03-130.9345700.0091170.0093841.0078120.0061040.1350100.0016040.007175NaN...NaNNaN0.0071870.0042340.005085NaN0.0058100.0029700.0085300
00000fd6641609c6ece5454664794f0340ad84dddce9a267a310b5ae68e9d8e52018-03-250.8803710.1781010.0346981.0039060.0069120.1655270.0055500.005070NaN...NaNNaN0.0029810.0074810.007874NaN0.0032840.0031700.0085140
00001b22f846c82c51f6e3958ccd81970162bae8b007e80662ef27519fcc18c12018-03-120.8808590.0097050.0042840.8125000.006451NaN0.0037960.007195NaN...NaNNaN0.0073810.0066220.000965NaN0.0022010.0008340.0034450
000041bdba6ecadd89a52d11886e8eaaec9325906c9723355abb5ca523658edc2018-03-290.6215820.0010820.0125661.0058590.0078280.2878420.0045320.009941NaN...NaNNaN0.0027050.0061840.001899NaN0.0081860.0055580.0029830
00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad51ca8b8c4a24cefed2018-03-300.8720700.0055730.0076790.8159180.001247NaN0.0002310.005527NaN...NaNNaN0.0029740.0041620.005764NaN0.0081560.0069430.0009050
\n

5 rows × 190 columns

\n
"},"metadata":{}}]},{"cell_type":"code","source":"train_dataset.info(max_cols=191,show_counts=True)","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:52:07.821671Z","iopub.execute_input":"2022-11-16T10:52:07.822452Z","iopub.status.idle":"2022-11-16T10:52:08.319794Z","shell.execute_reply.started":"2022-11-16T10:52:07.822409Z","shell.execute_reply":"2022-11-16T10:52:08.318519Z"},"trusted":true},"execution_count":7,"outputs":[{"name":"stdout","text":"\nIndex: 458913 entries, 0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fbac11a8ed792feb62a to fffff1d38b785cef84adeace64f8f83db3a0c31e8d92eaba8b115f71cab04681\nData columns (total 190 columns):\n # Column Non-Null Count Dtype \n--- ------ -------------- ----- \n 0 S_2 458913 non-null datetime64[ns]\n 1 P_2 455944 non-null float16 \n 2 D_39 458913 non-null float16 \n 3 B_1 458913 non-null float16 \n 4 B_2 458882 non-null float16 \n 5 R_1 458913 non-null float16 \n 6 S_3 373943 non-null float16 \n 7 D_41 458882 non-null float16 \n 8 B_3 458882 non-null float16 \n 9 D_42 59910 non-null float16 \n 10 D_43 324591 non-null float16 \n 11 D_44 436618 non-null float16 \n 12 B_4 458913 non-null float16 \n 13 D_45 458882 non-null float16 \n 14 B_5 458913 non-null float16 \n 15 R_2 458913 non-null float16 \n 16 D_46 363790 non-null float16 \n 17 D_47 458913 non-null float16 \n 18 D_48 400921 non-null float16 \n 19 D_49 51763 non-null float16 \n 20 B_6 458873 non-null float16 \n 21 B_7 458913 non-null float16 \n 22 B_8 454822 non-null float16 \n 23 D_50 196678 non-null float16 \n 24 D_51 458913 non-null float16 \n 25 B_9 458913 non-null float16 \n 26 R_3 458913 non-null float16 \n 27 D_52 457673 non-null float16 \n 28 P_3 436693 non-null float16 \n 29 B_10 458913 non-null float16 \n 30 D_53 132981 non-null float16 \n 31 S_5 458913 non-null float16 \n 32 B_11 458913 non-null float16 \n 33 S_6 458913 non-null float16 \n 34 D_54 458882 non-null float16 \n 35 R_4 458913 non-null float16 \n 36 S_7 373943 non-null float16 \n 37 B_12 458913 non-null float16 \n 38 S_8 458913 non-null float16 \n 39 D_55 428536 non-null float16 \n 40 D_56 214179 non-null float16 \n 41 B_13 457350 non-null float16 \n 42 R_5 458913 non-null float16 \n 43 D_58 458913 non-null float16 \n 44 S_9 275055 non-null float16 \n 45 B_14 458913 non-null float16 \n 46 D_59 454827 non-null float16 \n 47 D_60 458913 non-null float16 \n 48 D_61 410565 non-null float16 \n 49 B_15 458301 non-null float16 \n 50 S_11 458913 non-null float16 \n 51 D_62 399960 non-null float16 \n 52 D_63 458913 non-null category \n 53 D_64 458913 non-null category \n 54 D_65 458913 non-null float16 \n 55 B_16 458882 non-null float16 \n 56 B_17 214442 non-null float16 \n 57 B_18 458913 non-null float16 \n 58 B_19 458882 non-null float16 \n 59 D_66 52582 non-null category \n 60 B_20 458882 non-null float16 \n 61 D_68 449901 non-null category \n 62 S_12 458913 non-null float16 \n 63 R_6 458913 non-null float16 \n 64 S_13 458913 non-null float16 \n 65 B_21 458913 non-null float16 \n 66 D_69 452548 non-null float16 \n 67 B_22 458882 non-null float16 \n 68 D_70 454717 non-null float16 \n 69 D_71 458913 non-null float16 \n 70 D_72 457697 non-null float16 \n 71 S_15 458913 non-null float16 \n 72 B_23 458913 non-null float16 \n 73 D_73 4239 non-null float16 \n 74 P_4 458913 non-null float16 \n 75 D_74 457212 non-null float16 \n 76 D_75 458913 non-null float16 \n 77 D_76 49316 non-null float16 \n 78 B_24 458913 non-null float16 \n 79 R_7 458913 non-null float16 \n 80 D_77 245076 non-null float16 \n 81 B_25 458301 non-null float16 \n 82 B_26 458882 non-null float16 \n 83 D_78 436618 non-null float16 \n 84 D_79 456118 non-null float16 \n 85 R_8 458913 non-null float16 \n 86 R_9 26953 non-null float16 \n 87 S_16 458913 non-null float16 \n 88 D_80 457212 non-null float16 \n 89 R_10 458913 non-null float16 \n 90 R_11 458913 non-null float16 \n 91 B_27 458882 non-null float16 \n 92 D_81 457723 non-null float16 \n 93 D_82 115618 non-null float16 \n 94 S_17 458913 non-null float16 \n 95 R_12 458913 non-null float16 \n 96 B_28 458913 non-null float16 \n 97 R_13 458913 non-null float16 \n 98 D_83 452548 non-null float16 \n 99 R_14 458913 non-null float16 \n 100 R_15 458913 non-null float16 \n 101 D_84 457673 non-null float16 \n 102 R_16 458913 non-null float16 \n 103 B_29 27324 non-null float16 \n 104 B_30 458882 non-null category \n 105 S_18 458913 non-null float16 \n 106 D_86 458913 non-null float16 \n 107 D_87 645 non-null float16 \n 108 R_17 458913 non-null float16 \n 109 R_18 458913 non-null float16 \n 110 D_88 827 non-null float16 \n 111 B_31 458913 non-null float16 \n 112 S_19 458913 non-null float16 \n 113 R_19 458913 non-null float16 \n 114 B_32 458913 non-null float16 \n 115 S_20 458913 non-null float16 \n 116 R_20 458913 non-null float16 \n 117 R_21 458913 non-null float16 \n 118 B_33 458882 non-null float16 \n 119 D_89 457673 non-null float16 \n 120 R_22 458913 non-null float16 \n 121 R_23 458913 non-null float16 \n 122 D_91 446106 non-null float16 \n 123 D_92 458913 non-null float16 \n 124 D_93 458913 non-null float16 \n 125 D_94 458913 non-null float16 \n 126 R_24 458913 non-null float16 \n 127 R_25 458913 non-null float16 \n 128 D_96 458913 non-null float16 \n 129 S_22 457146 non-null float16 \n 130 S_23 458869 non-null float16 \n 131 S_24 457179 non-null float16 \n 132 S_25 457492 non-null float16 \n 133 S_26 458913 non-null float16 \n 134 D_102 458913 non-null float16 \n 135 D_103 456083 non-null float16 \n 136 D_104 456083 non-null float16 \n 137 D_105 213311 non-null float16 \n 138 D_106 51648 non-null float16 \n 139 D_107 456083 non-null float16 \n 140 B_36 458913 non-null float16 \n 141 B_37 458913 non-null float16 \n 142 R_26 51143 non-null float16 \n 143 R_27 430177 non-null float16 \n 144 B_38 458882 non-null category \n 145 D_108 2627 non-null float16 \n 146 D_109 458882 non-null float16 \n 147 D_110 3678 non-null float16 \n 148 D_111 3678 non-null float16 \n 149 B_39 4105 non-null float16 \n 150 D_112 458882 non-null float16 \n 151 B_40 458913 non-null float16 \n 152 S_27 341747 non-null float16 \n 153 D_113 452896 non-null float16 \n 154 D_114 452896 non-null category \n 155 D_115 452896 non-null float16 \n 156 D_116 452896 non-null category \n 157 D_117 452896 non-null category \n 158 D_118 452896 non-null float16 \n 159 D_119 452896 non-null float16 \n 160 D_120 452896 non-null category \n 161 D_121 452896 non-null float16 \n 162 D_122 452896 non-null float16 \n 163 D_123 452896 non-null float16 \n 164 D_124 452896 non-null float16 \n 165 D_125 452896 non-null float16 \n 166 D_126 458913 non-null category \n 167 D_127 458913 non-null float16 \n 168 D_128 456083 non-null float16 \n 169 D_129 456083 non-null float16 \n 170 B_41 458913 non-null float16 \n 171 B_42 6142 non-null float16 \n 172 D_130 456083 non-null float16 \n 173 D_131 456083 non-null float16 \n 174 D_132 51760 non-null float16 \n 175 D_133 458913 non-null float16 \n 176 R_28 458913 non-null float16 \n 177 D_134 16395 non-null float16 \n 178 D_135 16395 non-null float16 \n 179 D_136 16395 non-null float16 \n 180 D_137 16395 non-null float16 \n 181 D_138 16395 non-null float16 \n 182 D_139 456083 non-null float16 \n 183 D_140 458913 non-null float16 \n 184 D_141 456083 non-null float16 \n 185 D_142 80315 non-null float16 \n 186 D_143 456083 non-null float16 \n 187 D_144 458913 non-null float16 \n 188 D_145 456083 non-null float16 \n 189 target 458913 non-null int64 \ndtypes: category(11), datetime64[ns](1), float16(177), int64(1)\nmemory usage: 170.2+ MB\n","output_type":"stream"}]},{"cell_type":"code","source":"train_dataset.describe()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:52:13.163612Z","iopub.execute_input":"2022-11-16T10:52:13.16404Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Explore a Pattern","metadata":{}},{"cell_type":"markdown","source":"The dataset contains aggregated profile features for each customer at each statement date. Features are anonymized and normalized, and fall into the following general categories:\n\n* D_* = Delinquency variables (bad or criminal behaviour, especially among young people)\n* S_* = Spend variables\n* P_* = Payment variables\n* B_* = Balance variables\n* R_* = Risk variables\n\nwith the following features being categorical: ['B_30', 'B_38', 'D_114', 'D_116', 'D_117', 'D_120', 'D_126', 'D_63', 'D_64', 'D_66', 'D_68']","metadata":{}},{"cell_type":"code","source":"categorical_cols = ['B_30', 'B_38', 'D_114', 'D_116', 'D_117', 'D_120', 'D_126', 'D_63', 'D_64', 'D_66', 'D_68']\n\nnum_cols = [col for col in train_dataset.columns if col not in categorical_cols + [\"target\"]]\n\nprint(f'Total number of features: {1}')\nprint(f'Total number of categorical features: {len(categorical_cols)}')\nprint(f'Total number of continuos features: {len(num_cols)}')","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:23.835971Z","iopub.execute_input":"2022-11-16T10:06:23.836882Z","iopub.status.idle":"2022-11-16T10:06:23.845695Z","shell.execute_reply.started":"2022-11-16T10:06:23.83684Z","shell.execute_reply":"2022-11-16T10:06:23.843944Z"},"trusted":true},"execution_count":30,"outputs":[{"name":"stdout","text":"Total number of features: 1\nTotal number of categorical features: 11\nTotal number of continuos features: 178\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# Visualizing Target","metadata":{}},{"cell_type":"code","source":"sns.countplot(x = 'target', data = train_dataset)\n","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:23.848423Z","iopub.execute_input":"2022-11-16T10:06:23.849288Z","iopub.status.idle":"2022-11-16T10:06:24.205742Z","shell.execute_reply.started":"2022-11-16T10:06:23.84916Z","shell.execute_reply":"2022-11-16T10:06:24.204427Z"},"trusted":true},"execution_count":31,"outputs":[{"execution_count":31,"output_type":"execute_result","data":{"text/plain":""},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"# Visualizing categorial features","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(20, 30))\nfor i, k in enumerate(categorical_cols):\n plt.subplot(6, 2, i+1)\n temp_val = pd.DataFrame(train_dataset[k].value_counts(dropna=False, normalize=True).sort_index().rename('count'))\n temp_val.index.name = 'value'\n temp_val.reset_index(inplace=True)\n plt.bar(temp_val.index, temp_val['count'], alpha=0.5)\n plt.xlabel(k)\n plt.ylabel('frequency')\n plt.xticks(temp_val.index, temp_val.value)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:24.207291Z","iopub.execute_input":"2022-11-16T10:06:24.208506Z","iopub.status.idle":"2022-11-16T10:06:25.685663Z","shell.execute_reply.started":"2022-11-16T10:06:24.208457Z","shell.execute_reply":"2022-11-16T10:06:25.684042Z"},"trusted":true},"execution_count":32,"outputs":[{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"# Visualizing categorial features based on the target","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(20, 30))\nfor i, f in enumerate(categorical_cols):\n plt.subplot(6, 2, i+1)\n temp = pd.DataFrame(train_dataset[f][train_dataset.target == 0].value_counts(dropna=False, normalize=True).sort_index().rename('count'))\n temp.index.name = 'value'\n temp.reset_index(inplace=True)\n plt.bar(temp.index, temp['count'], alpha=0.5, label='target=0')\n temp = pd.DataFrame(train_dataset[f][train_dataset.target == 1].value_counts(dropna=False, normalize=True).sort_index().rename('count'))\n temp.index.name = 'value'\n temp.reset_index(inplace=True)\n plt.bar(temp.index, temp['count'], alpha=0.5, label='target=1')\n plt.xlabel(f)\n plt.ylabel('frequency')\n plt.legend()\n plt.xticks(temp.index, temp.value)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:25.687068Z","iopub.execute_input":"2022-11-16T10:06:25.687427Z","iopub.status.idle":"2022-11-16T10:06:27.867878Z","shell.execute_reply.started":"2022-11-16T10:06:25.687395Z","shell.execute_reply":"2022-11-16T10:06:27.866589Z"},"trusted":true},"execution_count":33,"outputs":[{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"# Aggregated profile features","metadata":{}},{"cell_type":"code","source":"Delinquency = [d for d in train_dataset.columns if d.startswith('D_')]\nSpend = [s for s in train_dataset.columns if s.startswith('S_')]\nPayment = [p for p in train_dataset.columns if p.startswith('P_')]\nBalance = [b for b in train_dataset.columns if b.startswith('B_')]\nRisk = [r for r in train_dataset.columns if r.startswith('R_')]\nDict = {'Delinquency': len(Delinquency), 'Spend': len(Spend), 'Payment': len(Payment), 'Balance': len(Balance), 'Risk': len(Risk),}\n\nplt.figure(figsize=(10,5))\nsns.barplot(x=list(Dict.keys()), y=list(Dict.values()));","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:27.869247Z","iopub.execute_input":"2022-11-16T10:06:27.869718Z","iopub.status.idle":"2022-11-16T10:06:28.075478Z","shell.execute_reply.started":"2022-11-16T10:06:27.869671Z","shell.execute_reply":"2022-11-16T10:06:28.074137Z"},"trusted":true},"execution_count":34,"outputs":[{"output_type":"display_data","data":{"text/plain":"
","image/png":"iVBORw0KGgoAAAANSUhEUgAAAlYAAAExCAYAAAC+vOxgAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAAVFUlEQVR4nO3dfbRlZ10f8O+PDIFAeAlmmhUScGIJ0IgQYUBsFCKhLFQo0UURijBVXGktBF+QgsVK9A9fKgKClpIGJLYYIQFXArIKaSQiiDETEggQkFkhQDCQwfIWoYSEX//Y+4abYWYyc89z59478/msddbZ789zznP2Od/z7H32qe4OAACLu9NaVwAA4GAhWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMcofBqqpeX1U3VtWHl027T1VdXFWfmO+PmqdXVb2qqnZU1Yeq6uGrWXkAgPVkX3qs3pDkibtMe3GSS7r7xCSXzONJ8qNJTpxvZyR5zZhqAgCsf7UvFwitqi1J3t7dD5nHP57k1O6+oaqOTXJpdz+oql47D5+363J72/7RRx/dW7ZsWeyRAAAcAFdcccUXunvz7uZtWuE2j1kWlj6X5Jh5+Lgkn1m23PXztL0Gqy1btmT79u0rrAoAwIFTVZ/a07yFT17vqctrv/8Xp6rOqKrtVbV9586di1YDAGDNrTRYfX4+BJj5/sZ5+meT3G/ZcsfP075Dd5/d3Vu7e+vmzbvtTQMA2FBWGqwuSrJtHt6W5MJl0589/zrw0Um+fEfnVwEAHCzu8ByrqjovyalJjq6q65O8NMnvJHlzVT0nyaeSPG1e/B1JfizJjiRfS/Izq1BnAIB16Q6DVXc/Yw+zTtvNsp3kuYtWCgBgI3LldQCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEFW+l+Ba+YRL/yTta7CIeGK33v2WlcBADYcPVYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgwhWAACDCFYAAIMIVgAAgywUrKrql6rqI1X14ao6r6ruWlUnVNVlVbWjqt5UVYePqiwAwHq24mBVVccleX6Srd39kCSHJXl6kt9N8orufkCSLyZ5zoiKAgCsd4seCtyU5Iiq2pTkbkluSPK4JBfM889NcvqCZQAAbAgrDlbd/dkkL0vy6UyB6stJrkjype6+ZV7s+iTHLVpJAICNYJFDgUcleUqSE5LcN8ndkzxxP9Y/o6q2V9X2nTt3rrQaAADrxiKHAh+f5JPdvbO7v5nkrUlOSXLv+dBgkhyf5LO7W7m7z+7urd29dfPmzQtUAwBgfVgkWH06yaOr6m5VVUlOS/LRJO9O8tR5mW1JLlysigAAG8Mi51hdlukk9Q8kuXre1tlJXpTkl6tqR5LvSvK6AfUEAFj3Nt3xInvW3S9N8tJdJl+b5FGLbBcAYCNy5XUAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBBCsAgEEEKwCAQQQrAIBBFgpWVXXvqrqgqj5WVddU1Q9W1X2q6uKq+sR8f9SoygIArGeL9lj9QZL/3d0PTvKwJNckeXGSS7r7xCSXzOMAAAe9FQerqrpXksckeV2SdPfN3f2lJE9Jcu682LlJTl+sigAAG8MiPVYnJNmZ5I+r6sqqOqeq7p7kmO6+YV7mc0mOWbSSAAAbwSLBalOShyd5TXd/f5J/yi6H/bq7k/TuVq6qM6pqe1Vt37lz5wLVAABYHxYJVtcnub67L5vHL8gUtD5fVccmyXx/4+5W7u6zu3trd2/dvHnzAtUAAFgfVhysuvtzST5TVQ+aJ52W5KNJLkqybZ62LcmFC9UQAGCD2LTg+mcmeWNVHZ7k2iQ/kymsvbmqnpPkU0metmAZAAAbwkLBqruvSrJ1N7NOW2S7AAAbkSuvAwAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADCJYAQAMIlgBAAwiWAEADLJwsKqqw6rqyqp6+zx+QlVdVlU7qupNVXX44tUEAFj/RvRY/UKSa5aN/26SV3T3A5J8MclzBpQBALDuLRSsqur4JD+e5Jx5vJI8LskF8yLnJjl9kTIAADaKRXusXpnkPyX51jz+XUm+1N23zOPXJzluwTIAADaETStdsaqelOTG7r6iqk5dwfpnJDkjSe5///uvtBoA7IO/esxj17oKB73Hvuev1roKrAOL9FidkuRfV9V1Sf4s0yHAP0hy76paCmzHJ/ns7lbu7rO7e2t3b928efMC1QAAWB9WHKy6+1e7+/ju3pLk6Un+srufmeTdSZ46L7YtyYUL1xIAYANYjetYvSjJL1fVjkznXL1uFcoAAFh3VnyO1XLdfWmSS+fha5M8asR2AQA2EldeBwAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGGTTWlcAANi7P3zB29a6Cge95/3+k4dsR48VAMAgghUAwCCCFQDAIIIVAMAgghUAwCCCFQDAIIIVAMAgghUAwCCCFQDAIIIVAMAgghUAwCCCFQDAIIIVAMAgghUAwCCCFQDAIJvWugIcWj79m9+31lU46N3/169e6yoAHLL0WAEADLLiYFVV96uqd1fVR6vqI1X1C/P0+1TVxVX1ifn+qHHVBQBYvxbpsbolyQu6+6Qkj07y3Ko6KcmLk1zS3ScmuWQeBwA46K04WHX3Dd39gXn4q0muSXJckqckOXde7Nwkpy9YRwCADWHIOVZVtSXJ9ye5LMkx3X3DPOtzSY4ZUQYAwHq3cLCqqiOTvCXJL3b3V5bP6+5O0ntY74yq2l5V23fu3LloNQAA1txCwaqq7pwpVL2xu986T/58VR07zz82yY27W7e7z+7urd29dfPmzYtUAwBgXVjkV4GV5HVJrunuly+bdVGSbfPwtiQXrrx6AAAbxyIXCD0lybOSXF1VV83T/nOS30ny5qp6TpJPJXnaQjUEANggVhysuvu9SWoPs09b6XYBADYqV14HABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGGTTWlcA2BhOefUpa12FQ8L7znzfWlcBWIAeKwCAQQQrAIBBBCsAgEFWJVhV1ROr6uNVtaOqXrwaZQAArDfDg1VVHZbkj5L8aJKTkjyjqk4aXQ4AwHqzGj1Wj0qyo7uv7e6bk/xZkqesQjkAAOvKagSr45J8Ztn49fM0AICDWnX32A1WPTXJE7v75+bxZyX5ge5+3i7LnZHkjHn0QUk+PrQi68vRSb6w1pVgRbTdxqb9Njbtt3Ed7G333d29eXczVuMCoZ9Ncr9l48fP026nu89OcvYqlL/uVNX27t661vVg/2m7jU37bWzab+M6lNtuNQ4FXp7kxKo6oaoOT/L0JBetQjkAAOvK8B6r7r6lqp6X5J1JDkvy+u7+yOhyAADWm1X5r8DufkeSd6zGtjeoQ+KQ50FK221s2m9j034b1yHbdsNPXgcAOFT5SxsAgEEOqWBVVbdW1VVV9ZGq+mBVvaCq9vocVNWWqvrwPLy1ql51YGrLSFX1krndPzS/Bn5gFcs6q6p+ZbW2f7Batn9+uKrOr6q7rXWdllTVqVX1L9e6HhvBsnb8YFV9YF+et6q66UDUjTF22VffVlX3nqfft6ou2Mt6t32eHswOqWCV5OvdfXJ3f2+Sf5Xpb3deuq8rd/f27n7+qtWOVVFVP5jkSUke3t0PTfL43P4itqwPS/vnQ5LcnOQ/rHWFljk1iWC1b5ba8WFJfjXJb691hRhu+b76f5M8N0m6+x+6+6lrW7W1d6gFq9t0942ZLlD6vJocVlW/V1WXz70a/37XdeZvrW+fh8+qqtdX1aVVdW1VPX/Zci+pqr+vqvdW1XlLvRfzslvn4aOr6rp5eLdlz+VdWlUXVNXHquqNVVXzvEdW1d/M3wr/rqruUVXvqaqTl9XjvVX1sNV6DjeQY5N8obu/kSTd/YXu/oequq6q/mtVXT0/hw9IkqraXFVvmdvj8qo6ZZ6+T22e6YK3LOavkzygqp5cVZdV1ZVV9X+q6piqulNVfaKqNifJPL5jbrc3VNVrqupv5zY6dW6za6rqDUsbr6onVNX75x6V86vqyHn6dVX1G/P0q6vqwVW1JVPI+6X5W/oPr8HzsVHdM8kXk6SqjqyqS5Y9t9/xV2d7Wmbu6bimqv5HTT3P76qqI+Z5D5hfGx+c1/vn8/QXLntP/Y0D+JgPNe/P/O8qdfsjPN87v69eNbfBictXqqrvmffrR65BnVdXdx8ytyQ37Wbal5Ickylk/do87S5Jtic5IcmWJB+ep5+a5O3z8FlJ/mZe9ugk/5jkzkkekeTqJHfL9KayI8mvzOtcmmTrPHx0kuvm4T2VfWqSL2e6yOqdMr2AfyjJ4UmuTfLIeZ17ZvqF57Ykr5ynPTDJ9rV+ztfDLcmRSa5K8vdJ/luSx87Tr0vyknn42cva9k+T/NA8fP8k16y0zd32q51umu83Jbkwyc8nOSrf/pHNzyX5/Xn4pUl+cR5+QpK3zMNvyPT/pJXpP0q/kuT75v3niiQnz233niR3n9d5UZJfX/aaOHMe/o9JzlnW9tp039rx1nl/+9j8/vWIZe16z3n46Hk/WWrbm/a2TKb34VuSnDzPe3OSn56HL0vyE/PwXef98AmZfpVWc9u/Pclj1vq5OVhuy9rrsCTnZ/q3leT2n5evTvLMefjwJEcszc/05fPKJA9b68eyGrdVudzCBvWEJA+t6S95kuReSU7M9GG8J3/RUy/IN6rqxkwB7YeT/Hl3fy1JqmpfLo66p7JvTvJ33X39vK2rMr0wv5zkhu6+PEm6+yvz/POT/JeqemGSn830IXPI6+6bquoRmdrmR5K8qapePM8+b9n9K+bhxyc5ae4cTJJ7LvVoZFyb852OmF/jydRj9bpMb8BvqqpjM705f3Ke//pM4euVmV7rf7xsO2/r7q6qq5N8vruvTpKq+kim/ef4JCcled/cxodn+tKy5K3z/RVJfnLcwztkfL27T05uOwz/J1X1kEwh57eq6jFJvpWpl+OYJJ9btu6elkmST3b3VfPwFUm2VNU9khzX3X+eJN39/+Zyn5DpffXKefkjM72nvmf4oz00Le2rxyW5JsnFu1nm/UleUlXHJ3lrd39i3t82Z9p3f7K7P3qA6ntAHdLBqqq+J9O3qxsz7dBndvc7d1lmy1428Y1lw7fmjp/PW/Ltw693XV7MHso+dX/K6O6vVdXFmb6pPy1TTwpJuvvWTD2Gl84fuNuWZi1fbL6/U5JHL71JL5nfFPa3zdl3t30gL6mqVyd5eXdfNO8PZyVJd3+mqj5fVY9L8qgkz1y22lIbfSu3b69vZWqvW5Nc3N3P2EM9ltbRvgvq7vdX1dGZPkx/bL5/RHd/s6ZTIe66yyrP3Msyu+57R+yl6Ery29392sUfBbvx9e4+uaYfmLwz0zlWt/thV3f/aVVdluTHk7yjplNcrs3UMfDpTEdfDspgdcieYzWfn/Hfk/xhT32V70zy81V153n+A6vq7ivY9HuSnF5VR8zfpp68bN51+XbYWX6C3/6W/fEkxy4dm67p/KqlD4BzMr3AL+/uL66g/gedqnrQLsf3T07yqXn4p5bdL/VavCvJmcvWP/kOithbm7OYe+Xb/zW6bZd55yT5X0nOn4PzvvrbJKfUt8+pu3tVPfAO1vlqknvsRxkkqaoHZzpc9I+Z2vLGOTD9SJLv3s0q+7LMbbr7q0mur6rT5/LusuzD/meXnTt3XFX9s1GPi8ncS//8JC9Y9hmU5LaOi2u7+1WZeqgeOs+6OclPJHl2Vf3bA1nfA+VQ+za21H1550y9R/8zycvneedkOkzwgZq6JnYmOX1/C+juD1TVm5J8MFNP2OXLZr8syZur6owkf7Fs+n6V3d03V9VPJXn1fALn1zMdvrqpu6+oqq/k9odGDnVHZnqu7p2p3XdkOq/tSUmOqqoPZfo2vNSD8fwkfzRP35QpOO3xF2p30OYs5qwk51fVF5P8ZaZzD5dclOl1vl+v9e7eWVX/Lsl5VXWXefKvZe+H/d+W5IL5ZOozu/uv96fMQ8zyQ7qVZFt331pVb0zytrnHeHumc7B2tS/L7OpZSV5bVb+Z5JtJ/k13v6uq/kWS9889zTcl+elM+ycDdfeV83vlMzIdwl/ytCTPqqpvZjrc+1uZzkFNd/9TVT0pycVVdVN3H1SnT7jy+iqrqrMyBZ6XHaDy7pvpkNeDu/tbB6LMjWo+zLC1u7+w1nVh/9X0C9tXdLdf6QHrxiF7KPBgVFXPzvQLmZcIVRzM5h8fvCXTdZIA1g09VgAAg+ixAgAYRLACABhEsAIAGESwAgAYRLACABhEsAIAGOT/A5r61RYruQSgAAAAAElFTkSuQmCC\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"# Check null values","metadata":{}},{"cell_type":"code","source":"NaN_Val = np.array(train_dataset.isnull().sum())\nNaN_prec = np.array((train_dataset.isnull().sum() * 100 / len(train_dataset)).round(2))\nNaN_Col = pd.DataFrame([np.array(list(train_dataset.columns)).T,NaN_Val.T,NaN_prec.T,np.array(list(train_dataset.dtypes)).T], index=['Features','Num of Missing values','Percentage','DataType']\n).transpose()\npd.set_option('display.max_rows', None)\nNaN_Col","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:28.08198Z","iopub.execute_input":"2022-11-16T10:06:28.082518Z","iopub.status.idle":"2022-11-16T10:06:29.04716Z","shell.execute_reply.started":"2022-11-16T10:06:28.082478Z","shell.execute_reply":"2022-11-16T10:06:29.045687Z"},"trusted":true},"execution_count":35,"outputs":[{"execution_count":35,"output_type":"execute_result","data":{"text/plain":" Features Num of Missing values Percentage DataType\n0 S_2 0 0.0 datetime64[ns]\n1 P_2 2969 0.65 float16\n2 D_39 0 0.0 float16\n3 B_1 0 0.0 float16\n4 B_2 31 0.01 float16\n5 R_1 0 0.0 float16\n6 S_3 84970 18.52 float16\n7 D_41 31 0.01 float16\n8 B_3 31 0.01 float16\n9 D_42 399003 86.95 float16\n10 D_43 134322 29.27 float16\n11 D_44 22295 4.86 float16\n12 B_4 0 0.0 float16\n13 D_45 31 0.01 float16\n14 B_5 0 0.0 float16\n15 R_2 0 0.0 float16\n16 D_46 95123 20.73 float16\n17 D_47 0 0.0 float16\n18 D_48 57992 12.64 float16\n19 D_49 407150 88.72 float16\n20 B_6 40 0.01 float16\n21 B_7 0 0.0 float16\n22 B_8 4091 0.89 float16\n23 D_50 262235 57.14 float16\n24 D_51 0 0.0 float16\n25 B_9 0 0.0 float16\n26 R_3 0 0.0 float16\n27 D_52 1240 0.27 float16\n28 P_3 22220 4.84 float16\n29 B_10 0 0.0 float16\n30 D_53 325932 71.02 float16\n31 S_5 0 0.0 float16\n32 B_11 0 0.0 float16\n33 S_6 0 0.0 float16\n34 D_54 31 0.01 float16\n35 R_4 0 0.0 float16\n36 S_7 84970 18.52 float16\n37 B_12 0 0.0 float16\n38 S_8 0 0.0 float16\n39 D_55 30377 6.62 float16\n40 D_56 244734 53.33 float16\n41 B_13 1563 0.34 float16\n42 R_5 0 0.0 float16\n43 D_58 0 0.0 float16\n44 S_9 183858 40.06 float16\n45 B_14 0 0.0 float16\n46 D_59 4086 0.89 float16\n47 D_60 0 0.0 float16\n48 D_61 48348 10.54 float16\n49 B_15 612 0.13 float16\n50 S_11 0 0.0 float16\n51 D_62 58953 12.85 float16\n52 D_63 0 0.0 category\n53 D_64 0 0.0 category\n54 D_65 0 0.0 float16\n55 B_16 31 0.01 float16\n56 B_17 244471 53.27 float16\n57 B_18 0 0.0 float16\n58 B_19 31 0.01 float16\n59 D_66 406331 88.54 category\n60 B_20 31 0.01 float16\n61 D_68 9012 1.96 category\n62 S_12 0 0.0 float16\n63 R_6 0 0.0 float16\n64 S_13 0 0.0 float16\n65 B_21 0 0.0 float16\n66 D_69 6365 1.39 float16\n67 B_22 31 0.01 float16\n68 D_70 4196 0.91 float16\n69 D_71 0 0.0 float16\n70 D_72 1216 0.26 float16\n71 S_15 0 0.0 float16\n72 B_23 0 0.0 float16\n73 D_73 454674 99.08 float16\n74 P_4 0 0.0 float16\n75 D_74 1701 0.37 float16\n76 D_75 0 0.0 float16\n77 D_76 409597 89.25 float16\n78 B_24 0 0.0 float16\n79 R_7 0 0.0 float16\n80 D_77 213837 46.6 float16\n81 B_25 612 0.13 float16\n82 B_26 31 0.01 float16\n83 D_78 22295 4.86 float16\n84 D_79 2795 0.61 float16\n85 R_8 0 0.0 float16\n86 R_9 431960 94.13 float16\n87 S_16 0 0.0 float16\n88 D_80 1701 0.37 float16\n89 R_10 0 0.0 float16\n90 R_11 0 0.0 float16\n91 B_27 31 0.01 float16\n92 D_81 1190 0.26 float16\n93 D_82 343295 74.81 float16\n94 S_17 0 0.0 float16\n95 R_12 0 0.0 float16\n96 B_28 0 0.0 float16\n97 R_13 0 0.0 float16\n98 D_83 6365 1.39 float16\n99 R_14 0 0.0 float16\n100 R_15 0 0.0 float16\n101 D_84 1240 0.27 float16\n102 R_16 0 0.0 float16\n103 B_29 431589 94.05 float16\n104 B_30 31 0.01 category\n105 S_18 0 0.0 float16\n106 D_86 0 0.0 float16\n107 D_87 458268 99.86 float16\n108 R_17 0 0.0 float16\n109 R_18 0 0.0 float16\n110 D_88 458086 99.82 float16\n111 B_31 0 0.0 float16\n112 S_19 0 0.0 float16\n113 R_19 0 0.0 float16\n114 B_32 0 0.0 float16\n115 S_20 0 0.0 float16\n116 R_20 0 0.0 float16\n117 R_21 0 0.0 float16\n118 B_33 31 0.01 float16\n119 D_89 1240 0.27 float16\n120 R_22 0 0.0 float16\n121 R_23 0 0.0 float16\n122 D_91 12807 2.79 float16\n123 D_92 0 0.0 float16\n124 D_93 0 0.0 float16\n125 D_94 0 0.0 float16\n126 R_24 0 0.0 float16\n127 R_25 0 0.0 float16\n128 D_96 0 0.0 float16\n129 S_22 1767 0.39 float16\n130 S_23 44 0.01 float16\n131 S_24 1734 0.38 float16\n132 S_25 1421 0.31 float16\n133 S_26 0 0.0 float16\n134 D_102 0 0.0 float16\n135 D_103 2830 0.62 float16\n136 D_104 2830 0.62 float16\n137 D_105 245602 53.52 float16\n138 D_106 407265 88.75 float16\n139 D_107 2830 0.62 float16\n140 B_36 0 0.0 float16\n141 B_37 0 0.0 float16\n142 R_26 407770 88.86 float16\n143 R_27 28736 6.26 float16\n144 B_38 31 0.01 category\n145 D_108 456286 99.43 float16\n146 D_109 31 0.01 float16\n147 D_110 455235 99.2 float16\n148 D_111 455235 99.2 float16\n149 B_39 454808 99.11 float16\n150 D_112 31 0.01 float16\n151 B_40 0 0.0 float16\n152 S_27 117166 25.53 float16\n153 D_113 6017 1.31 float16\n154 D_114 6017 1.31 category\n155 D_115 6017 1.31 float16\n156 D_116 6017 1.31 category\n157 D_117 6017 1.31 category\n158 D_118 6017 1.31 float16\n159 D_119 6017 1.31 float16\n160 D_120 6017 1.31 category\n161 D_121 6017 1.31 float16\n162 D_122 6017 1.31 float16\n163 D_123 6017 1.31 float16\n164 D_124 6017 1.31 float16\n165 D_125 6017 1.31 float16\n166 D_126 0 0.0 category\n167 D_127 0 0.0 float16\n168 D_128 2830 0.62 float16\n169 D_129 2830 0.62 float16\n170 B_41 0 0.0 float16\n171 B_42 452771 98.66 float16\n172 D_130 2830 0.62 float16\n173 D_131 2830 0.62 float16\n174 D_132 407153 88.72 float16\n175 D_133 0 0.0 float16\n176 R_28 0 0.0 float16\n177 D_134 442518 96.43 float16\n178 D_135 442518 96.43 float16\n179 D_136 442518 96.43 float16\n180 D_137 442518 96.43 float16\n181 D_138 442518 96.43 float16\n182 D_139 2830 0.62 float16\n183 D_140 0 0.0 float16\n184 D_141 2830 0.62 float16\n185 D_142 378598 82.5 float16\n186 D_143 2830 0.62 float16\n187 D_144 0 0.0 float16\n188 D_145 2830 0.62 float16\n189 target 0 0.0 int64","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
FeaturesNum of Missing valuesPercentageDataType
0S_200.0datetime64[ns]
1P_229690.65float16
2D_3900.0float16
3B_100.0float16
4B_2310.01float16
5R_100.0float16
6S_38497018.52float16
7D_41310.01float16
8B_3310.01float16
9D_4239900386.95float16
10D_4313432229.27float16
11D_44222954.86float16
12B_400.0float16
13D_45310.01float16
14B_500.0float16
15R_200.0float16
16D_469512320.73float16
17D_4700.0float16
18D_485799212.64float16
19D_4940715088.72float16
20B_6400.01float16
21B_700.0float16
22B_840910.89float16
23D_5026223557.14float16
24D_5100.0float16
25B_900.0float16
26R_300.0float16
27D_5212400.27float16
28P_3222204.84float16
29B_1000.0float16
30D_5332593271.02float16
31S_500.0float16
32B_1100.0float16
33S_600.0float16
34D_54310.01float16
35R_400.0float16
36S_78497018.52float16
37B_1200.0float16
38S_800.0float16
39D_55303776.62float16
40D_5624473453.33float16
41B_1315630.34float16
42R_500.0float16
43D_5800.0float16
44S_918385840.06float16
45B_1400.0float16
46D_5940860.89float16
47D_6000.0float16
48D_614834810.54float16
49B_156120.13float16
50S_1100.0float16
51D_625895312.85float16
52D_6300.0category
53D_6400.0category
54D_6500.0float16
55B_16310.01float16
56B_1724447153.27float16
57B_1800.0float16
58B_19310.01float16
59D_6640633188.54category
60B_20310.01float16
61D_6890121.96category
62S_1200.0float16
63R_600.0float16
64S_1300.0float16
65B_2100.0float16
66D_6963651.39float16
67B_22310.01float16
68D_7041960.91float16
69D_7100.0float16
70D_7212160.26float16
71S_1500.0float16
72B_2300.0float16
73D_7345467499.08float16
74P_400.0float16
75D_7417010.37float16
76D_7500.0float16
77D_7640959789.25float16
78B_2400.0float16
79R_700.0float16
80D_7721383746.6float16
81B_256120.13float16
82B_26310.01float16
83D_78222954.86float16
84D_7927950.61float16
85R_800.0float16
86R_943196094.13float16
87S_1600.0float16
88D_8017010.37float16
89R_1000.0float16
90R_1100.0float16
91B_27310.01float16
92D_8111900.26float16
93D_8234329574.81float16
94S_1700.0float16
95R_1200.0float16
96B_2800.0float16
97R_1300.0float16
98D_8363651.39float16
99R_1400.0float16
100R_1500.0float16
101D_8412400.27float16
102R_1600.0float16
103B_2943158994.05float16
104B_30310.01category
105S_1800.0float16
106D_8600.0float16
107D_8745826899.86float16
108R_1700.0float16
109R_1800.0float16
110D_8845808699.82float16
111B_3100.0float16
112S_1900.0float16
113R_1900.0float16
114B_3200.0float16
115S_2000.0float16
116R_2000.0float16
117R_2100.0float16
118B_33310.01float16
119D_8912400.27float16
120R_2200.0float16
121R_2300.0float16
122D_91128072.79float16
123D_9200.0float16
124D_9300.0float16
125D_9400.0float16
126R_2400.0float16
127R_2500.0float16
128D_9600.0float16
129S_2217670.39float16
130S_23440.01float16
131S_2417340.38float16
132S_2514210.31float16
133S_2600.0float16
134D_10200.0float16
135D_10328300.62float16
136D_10428300.62float16
137D_10524560253.52float16
138D_10640726588.75float16
139D_10728300.62float16
140B_3600.0float16
141B_3700.0float16
142R_2640777088.86float16
143R_27287366.26float16
144B_38310.01category
145D_10845628699.43float16
146D_109310.01float16
147D_11045523599.2float16
148D_11145523599.2float16
149B_3945480899.11float16
150D_112310.01float16
151B_4000.0float16
152S_2711716625.53float16
153D_11360171.31float16
154D_11460171.31category
155D_11560171.31float16
156D_11660171.31category
157D_11760171.31category
158D_11860171.31float16
159D_11960171.31float16
160D_12060171.31category
161D_12160171.31float16
162D_12260171.31float16
163D_12360171.31float16
164D_12460171.31float16
165D_12560171.31float16
166D_12600.0category
167D_12700.0float16
168D_12828300.62float16
169D_12928300.62float16
170B_4100.0float16
171B_4245277198.66float16
172D_13028300.62float16
173D_13128300.62float16
174D_13240715388.72float16
175D_13300.0float16
176R_2800.0float16
177D_13444251896.43float16
178D_13544251896.43float16
179D_13644251896.43float16
180D_13744251896.43float16
181D_13844251896.43float16
182D_13928300.62float16
183D_14000.0float16
184D_14128300.62float16
185D_14237859882.5float16
186D_14328300.62float16
187D_14400.0float16
188D_14528300.62float16
189target00.0int64
\n
"},"metadata":{}}]},{"cell_type":"markdown","source":"There are many missing values in the dataset","metadata":{}},{"cell_type":"markdown","source":"# Drop unuseful columns","metadata":{}},{"cell_type":"markdown","source":"Remove columns if there are > 80% of missing values","metadata":{}},{"cell_type":"code","source":"train_dataset = train_dataset.drop(['S_2','D_66','D_42','D_49','D_73','D_76','R_9','B_29','D_87','D_88','D_106','R_26','D_108','D_110','D_111','B_39','B_42','D_132','D_134','D_135','D_136','D_137','D_138','D_142'], axis=1)","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:29.049839Z","iopub.execute_input":"2022-11-16T10:06:29.050696Z","iopub.status.idle":"2022-11-16T10:06:29.367129Z","shell.execute_reply.started":"2022-11-16T10:06:29.050655Z","shell.execute_reply":"2022-11-16T10:06:29.365316Z"},"trusted":true},"execution_count":36,"outputs":[]},{"cell_type":"markdown","source":"# Fill null values","metadata":{}},{"cell_type":"code","source":"selected_col = np.array(['P_2','S_3','B_2','D_41','D_43','B_3','D_44','D_45','D_46','D_48','D_50','D_53','S_7','D_56','S_9','B_6','B_8','D_52','P_3','D_54','D_55','B_13','D_59','D_61','B_15','D_62','B_16','B_17','D_77','B_19','B_20','D_69','B_22','D_70','D_72','D_74','R_7','B_25','B_26','D_78','D_79','D_80','B_27','D_81','R_12','D_82','D_105','S_27','D_83','R_14','D_84','D_86','R_20','B_33','D_89','D_91','S_22','S_23','S_24','S_25','S_26','D_102','D_103','D_104','D_107','B_37','R_27','D_109','D_112','B_40','D_113','D_115','D_118','D_119','D_121','D_122','D_123','D_124','D_125','D_128','D_129','B_41','D_130','D_131','D_133','D_139','D_140','D_141','D_143','D_144','D_145'])\n\nfor col in selected_col:\n train_dataset[col] = train_dataset[col].fillna(train_dataset[col].median())","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:29.369905Z","iopub.execute_input":"2022-11-16T10:06:29.370792Z","iopub.status.idle":"2022-11-16T10:06:31.296401Z","shell.execute_reply.started":"2022-11-16T10:06:29.370742Z","shell.execute_reply":"2022-11-16T10:06:31.294568Z"},"trusted":true},"execution_count":37,"outputs":[]},{"cell_type":"markdown","source":"In describe session you saw, lot of cloumns means are NaN. So, that's why i have used median to fill NaN values. ","metadata":{}},{"cell_type":"code","source":"selcted_col2 = np.array(['D_68','B_30','B_38','D_64','D_114','D_116','D_117','D_120','D_126'])\n\nfor col2 in selcted_col2:\n train_dataset[col2] = train_dataset[col2].fillna(train_dataset[col2].mode()[0])","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:31.29895Z","iopub.execute_input":"2022-11-16T10:06:31.299572Z","iopub.status.idle":"2022-11-16T10:06:31.362398Z","shell.execute_reply.started":"2022-11-16T10:06:31.299518Z","shell.execute_reply":"2022-11-16T10:06:31.360638Z"},"trusted":true},"execution_count":38,"outputs":[]},{"cell_type":"markdown","source":"# Check again null values","metadata":{}},{"cell_type":"code","source":"print(train_dataset.isnull().sum().to_string())","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:31.364376Z","iopub.execute_input":"2022-11-16T10:06:31.364809Z","iopub.status.idle":"2022-11-16T10:06:31.733057Z","shell.execute_reply.started":"2022-11-16T10:06:31.364776Z","shell.execute_reply":"2022-11-16T10:06:31.731741Z"},"trusted":true},"execution_count":39,"outputs":[{"name":"stdout","text":"P_2 0\nD_39 0\nB_1 0\nB_2 0\nR_1 0\nS_3 0\nD_41 0\nB_3 0\nD_43 0\nD_44 0\nB_4 0\nD_45 0\nB_5 0\nR_2 0\nD_46 0\nD_47 0\nD_48 0\nB_6 0\nB_7 0\nB_8 0\nD_50 0\nD_51 0\nB_9 0\nR_3 0\nD_52 0\nP_3 0\nB_10 0\nD_53 0\nS_5 0\nB_11 0\nS_6 0\nD_54 0\nR_4 0\nS_7 0\nB_12 0\nS_8 0\nD_55 0\nD_56 0\nB_13 0\nR_5 0\nD_58 0\nS_9 0\nB_14 0\nD_59 0\nD_60 0\nD_61 0\nB_15 0\nS_11 0\nD_62 0\nD_63 0\nD_64 0\nD_65 0\nB_16 0\nB_17 0\nB_18 0\nB_19 0\nB_20 0\nD_68 0\nS_12 0\nR_6 0\nS_13 0\nB_21 0\nD_69 0\nB_22 0\nD_70 0\nD_71 0\nD_72 0\nS_15 0\nB_23 0\nP_4 0\nD_74 0\nD_75 0\nB_24 0\nR_7 0\nD_77 0\nB_25 0\nB_26 0\nD_78 0\nD_79 0\nR_8 0\nS_16 0\nD_80 0\nR_10 0\nR_11 0\nB_27 0\nD_81 0\nD_82 0\nS_17 0\nR_12 0\nB_28 0\nR_13 0\nD_83 0\nR_14 0\nR_15 0\nD_84 0\nR_16 0\nB_30 0\nS_18 0\nD_86 0\nR_17 0\nR_18 0\nB_31 0\nS_19 0\nR_19 0\nB_32 0\nS_20 0\nR_20 0\nR_21 0\nB_33 0\nD_89 0\nR_22 0\nR_23 0\nD_91 0\nD_92 0\nD_93 0\nD_94 0\nR_24 0\nR_25 0\nD_96 0\nS_22 0\nS_23 0\nS_24 0\nS_25 0\nS_26 0\nD_102 0\nD_103 0\nD_104 0\nD_105 0\nD_107 0\nB_36 0\nB_37 0\nR_27 0\nB_38 0\nD_109 0\nD_112 0\nB_40 0\nS_27 0\nD_113 0\nD_114 0\nD_115 0\nD_116 0\nD_117 0\nD_118 0\nD_119 0\nD_120 0\nD_121 0\nD_122 0\nD_123 0\nD_124 0\nD_125 0\nD_126 0\nD_127 0\nD_128 0\nD_129 0\nB_41 0\nD_130 0\nD_131 0\nD_133 0\nR_28 0\nD_139 0\nD_140 0\nD_141 0\nD_143 0\nD_144 0\nD_145 0\ntarget 0\n","output_type":"stream"}]},{"cell_type":"markdown","source":"There are no more missing values","metadata":{}},{"cell_type":"code","source":"train_dataset.shape","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:31.734666Z","iopub.execute_input":"2022-11-16T10:06:31.735025Z","iopub.status.idle":"2022-11-16T10:06:31.741973Z","shell.execute_reply.started":"2022-11-16T10:06:31.734994Z","shell.execute_reply":"2022-11-16T10:06:31.740754Z"},"trusted":true},"execution_count":40,"outputs":[{"execution_count":40,"output_type":"execute_result","data":{"text/plain":"(458913, 166)"},"metadata":{}}]},{"cell_type":"code","source":"train_dataset.head()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:31.74382Z","iopub.execute_input":"2022-11-16T10:06:31.744231Z","iopub.status.idle":"2022-11-16T10:06:31.781323Z","shell.execute_reply.started":"2022-11-16T10:06:31.744197Z","shell.execute_reply":"2022-11-16T10:06:31.779977Z"},"trusted":true},"execution_count":41,"outputs":[{"execution_count":41,"output_type":"execute_result","data":{"text/plain":" P_2 D_39 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.934570 0.009117 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.880371 0.178101 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.880859 0.009705 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.621582 0.001082 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.872070 0.005573 \n\n B_1 B_2 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.009384 1.007812 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.034698 1.003906 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.004284 0.812500 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.012566 1.005859 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.007679 0.815918 \n\n R_1 S_3 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.006104 0.135010 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.006912 0.165527 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.006451 0.164917 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.007828 0.287842 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.001247 0.164917 \n\n D_41 B_3 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.001604 0.007175 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.005550 0.005070 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.003796 0.007195 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.004532 0.009941 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.000231 0.005527 \n\n D_43 D_44 ... \\\ncustomer_ID ... \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.092407 0.003258 ... \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.060638 0.008781 ... \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.092407 0.000628 ... \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.046112 0.007793 ... \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.044678 0.002436 ... \n\n D_131 D_133 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.005703 0.006210 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.001928 0.002996 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.003469 0.009880 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.004578 0.001789 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.008896 0.005047 \n\n R_28 D_139 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.002716 0.007187 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.001701 0.002981 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.007690 0.007381 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.005138 0.002705 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.003706 0.002974 \n\n D_140 D_141 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.004234 0.005085 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.007481 0.007874 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.006622 0.000965 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.006184 0.001899 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.004162 0.005764 \n\n D_143 D_144 \\\ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.005810 0.002970 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.003284 0.003170 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.002201 0.000834 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.008186 0.005558 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.008156 0.006943 \n\n D_145 target \ncustomer_ID \n0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fb... 0.008530 0 \n00000fd6641609c6ece5454664794f0340ad84dddce9a26... 0.008514 0 \n00001b22f846c82c51f6e3958ccd81970162bae8b007e80... 0.003445 0 \n000041bdba6ecadd89a52d11886e8eaaec9325906c97233... 0.002983 0 \n00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad... 0.000905 0 \n\n[5 rows x 166 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
P_2D_39B_1B_2R_1S_3D_41B_3D_43D_44...D_131D_133R_28D_139D_140D_141D_143D_144D_145target
customer_ID
0000099d6bd597052cdcda90ffabf56573fe9d7c79be5fbac11a8ed792feb62a0.9345700.0091170.0093841.0078120.0061040.1350100.0016040.0071750.0924070.003258...0.0057030.0062100.0027160.0071870.0042340.0050850.0058100.0029700.0085300
00000fd6641609c6ece5454664794f0340ad84dddce9a267a310b5ae68e9d8e50.8803710.1781010.0346981.0039060.0069120.1655270.0055500.0050700.0606380.008781...0.0019280.0029960.0017010.0029810.0074810.0078740.0032840.0031700.0085140
00001b22f846c82c51f6e3958ccd81970162bae8b007e80662ef27519fcc18c10.8808590.0097050.0042840.8125000.0064510.1649170.0037960.0071950.0924070.000628...0.0034690.0098800.0076900.0073810.0066220.0009650.0022010.0008340.0034450
000041bdba6ecadd89a52d11886e8eaaec9325906c9723355abb5ca523658edc0.6215820.0010820.0125661.0058590.0078280.2878420.0045320.0099410.0461120.007793...0.0045780.0017890.0051380.0027050.0061840.0018990.0081860.0055580.0029830
00007889e4fcd2614b6cbe7f8f3d2e5c728eca32d9eb8ad51ca8b8c4a24cefed0.8720700.0055730.0076790.8159180.0012470.1649170.0002310.0055270.0446780.002436...0.0088960.0050470.0037060.0029740.0041620.0057640.0081560.0069430.0009050
\n

5 rows × 166 columns

\n
"},"metadata":{}}]},{"cell_type":"markdown","source":"# Load Testing DataSet","metadata":{}},{"cell_type":"code","source":"test_dataset_ = pd.read_feather('../input/amexfeather/test_data.ftr')\n# Keep the latest statement features for each customer\ntest_dataset = test_dataset_.groupby('customer_ID').tail(1).set_index('customer_ID', drop=True).sort_index()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:06:31.783211Z","iopub.execute_input":"2022-11-16T10:06:31.784233Z","iopub.status.idle":"2022-11-16T10:07:19.849615Z","shell.execute_reply.started":"2022-11-16T10:06:31.784192Z","shell.execute_reply":"2022-11-16T10:07:19.844099Z"},"trusted":true},"execution_count":42,"outputs":[]},{"cell_type":"code","source":"del test_dataset_\ngc.collect()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:19.855249Z","iopub.execute_input":"2022-11-16T10:07:19.855937Z","iopub.status.idle":"2022-11-16T10:07:20.515676Z","shell.execute_reply.started":"2022-11-16T10:07:19.855893Z","shell.execute_reply":"2022-11-16T10:07:20.51351Z"},"trusted":true},"execution_count":43,"outputs":[{"execution_count":43,"output_type":"execute_result","data":{"text/plain":"71869"},"metadata":{}}]},{"cell_type":"code","source":"test_dataset.head()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:20.517811Z","iopub.execute_input":"2022-11-16T10:07:20.518351Z","iopub.status.idle":"2022-11-16T10:07:20.562359Z","shell.execute_reply.started":"2022-11-16T10:07:20.518315Z","shell.execute_reply":"2022-11-16T10:07:20.560806Z"},"trusted":true},"execution_count":44,"outputs":[{"execution_count":44,"output_type":"execute_result","data":{"text/plain":" S_2 P_2 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 2019-10-12 0.568848 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 2019-04-15 0.841309 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 2019-10-16 0.697754 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 2019-04-22 0.513184 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 2019-10-22 0.254395 \n\n D_39 B_1 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.121399 0.010780 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.126465 0.016556 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.002724 0.001485 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.324707 0.149536 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.768066 0.563477 \n\n B_2 R_1 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 1.009766 0.006924 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 1.008789 0.009712 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.810059 0.002621 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.205688 0.002277 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.038025 0.502930 \n\n S_3 D_41 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.149414 0.000396 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.112183 0.006191 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.166138 0.004887 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.181152 0.005814 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.168335 0.009483 \n\n B_3 D_42 ... \\\ncustomer_ID ... \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.003576 0.10376 ... \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.011383 NaN ... \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.015945 NaN ... \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.498535 NaN ... \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.831055 NaN ... \n\n D_136 D_137 D_138 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... NaN NaN NaN \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... NaN NaN NaN \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... NaN NaN NaN \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... NaN NaN NaN \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... NaN NaN NaN \n\n D_139 D_140 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.005913 0.001250 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.004345 0.000866 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 1.000977 0.008896 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 1.007812 0.003754 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.006622 0.001140 \n\n D_141 D_142 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.006542 NaN \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.009117 NaN \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.895996 0.150146 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.919922 0.255371 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.009529 NaN \n\n D_143 D_144 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.009163 0.003691 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.002197 0.000247 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 1.009766 0.457764 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 1.007812 0.500977 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.009407 0.001557 \n\n D_145 \ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.003220 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.007778 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.092041 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.182983 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.000525 \n\n[5 rows x 189 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
S_2P_2D_39B_1B_2R_1S_3D_41B_3D_42...D_136D_137D_138D_139D_140D_141D_142D_143D_144D_145
customer_ID
00000469ba478561f23a92a868bd366de6f6527a684c9a2e78fb826dcac3b9b72019-10-120.5688480.1213990.0107801.0097660.0069240.1494140.0003960.0035760.10376...NaNNaNNaN0.0059130.0012500.006542NaN0.0091630.0036910.003220
00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397d4263dafa1daedef52019-04-150.8413090.1264650.0165561.0087890.0097120.1121830.0061910.011383NaN...NaNNaNNaN0.0043450.0008660.009117NaN0.0021970.0002470.007778
0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5e400fc98e7bd43ce82019-10-160.6977540.0027240.0014850.8100590.0026210.1661380.0048870.015945NaN...NaNNaNNaN1.0009770.0088960.8959960.1501461.0097660.4577640.092041
00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf6e56734528702d6942019-04-220.5131840.3247070.1495360.2056880.0022770.1811520.0058140.498535NaN...NaNNaNNaN1.0078120.0037540.9199220.2553711.0078120.5009770.182983
00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a4693dd914fca225572019-10-220.2543950.7680660.5634770.0380250.5029300.1683350.0094830.831055NaN...NaNNaNNaN0.0066220.0011400.009529NaN0.0094070.0015570.000525
\n

5 rows × 189 columns

\n
"},"metadata":{}}]},{"cell_type":"code","source":"test_dataset.shape","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:20.564194Z","iopub.execute_input":"2022-11-16T10:07:20.564679Z","iopub.status.idle":"2022-11-16T10:07:20.578799Z","shell.execute_reply.started":"2022-11-16T10:07:20.564641Z","shell.execute_reply":"2022-11-16T10:07:20.577524Z"},"trusted":true},"execution_count":45,"outputs":[{"execution_count":45,"output_type":"execute_result","data":{"text/plain":"(924621, 189)"},"metadata":{}}]},{"cell_type":"markdown","source":"# Check null values","metadata":{}},{"cell_type":"code","source":"NaN_Val2 = np.array(test_dataset.isnull().sum())\nNaN_prec2 = np.array((test_dataset.isnull().sum() * 100 / len(test_dataset)).round(2))\nNaN_Col2 = pd.DataFrame([np.array(list(test_dataset.columns)).T,NaN_Val2.T,NaN_prec2.T,np.array(list(test_dataset.dtypes)).T], index=['Features','Num of Missing values','Percentage','DataType']\n).transpose()\npd.set_option('display.max_rows', None)\n\nNaN_Col2","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:20.580574Z","iopub.execute_input":"2022-11-16T10:07:20.581401Z","iopub.status.idle":"2022-11-16T10:07:22.480766Z","shell.execute_reply.started":"2022-11-16T10:07:20.581349Z","shell.execute_reply":"2022-11-16T10:07:22.47922Z"},"trusted":true},"execution_count":46,"outputs":[{"execution_count":46,"output_type":"execute_result","data":{"text/plain":" Features Num of Missing values Percentage DataType\n0 S_2 0 0.0 datetime64[ns]\n1 P_2 4784 0.52 float16\n2 D_39 0 0.0 float16\n3 B_1 0 0.0 float16\n4 B_2 43 0.0 float16\n5 R_1 0 0.0 float16\n6 S_3 144280 15.6 float16\n7 D_41 43 0.0 float16\n8 B_3 43 0.0 float16\n9 D_42 827578 89.5 float16\n10 D_43 272218 29.44 float16\n11 D_44 44612 4.82 float16\n12 B_4 0 0.0 float16\n13 D_45 45 0.0 float16\n14 B_5 0 0.0 float16\n15 R_2 0 0.0 float16\n16 D_46 179260 19.39 float16\n17 D_47 0 0.0 float16\n18 D_48 116986 12.65 float16\n19 D_49 802639 86.81 float16\n20 B_6 138 0.01 float16\n21 B_7 0 0.0 float16\n22 B_8 12384 1.34 float16\n23 D_50 529868 57.31 float16\n24 D_51 0 0.0 float16\n25 B_9 0 0.0 float16\n26 R_3 0 0.0 float16\n27 D_52 2209 0.24 float16\n28 P_3 30460 3.29 float16\n29 B_10 0 0.0 float16\n30 D_53 661023 71.49 float16\n31 S_5 0 0.0 float16\n32 B_11 0 0.0 float16\n33 S_6 0 0.0 float16\n34 D_54 43 0.0 float16\n35 R_4 0 0.0 float16\n36 S_7 144280 15.6 float16\n37 B_12 0 0.0 float16\n38 S_8 0 0.0 float16\n39 D_55 60438 6.54 float16\n40 D_56 463223 50.1 float16\n41 B_13 2010 0.22 float16\n42 R_5 0 0.0 float16\n43 D_58 0 0.0 float16\n44 S_9 200608 21.7 float16\n45 B_14 0 0.0 float16\n46 D_59 8991 0.97 float16\n47 D_60 0 0.0 float16\n48 D_61 101585 10.99 float16\n49 B_15 1409 0.15 float16\n50 S_11 0 0.0 float16\n51 D_62 154713 16.73 float16\n52 D_63 0 0.0 category\n53 D_64 0 0.0 category\n54 D_65 0 0.0 float16\n55 B_16 43 0.0 float16\n56 B_17 482036 52.13 float16\n57 B_18 0 0.0 float16\n58 B_19 43 0.0 float16\n59 D_66 813168 87.95 category\n60 B_20 43 0.0 float16\n61 D_68 22380 2.42 category\n62 S_12 611 0.07 float16\n63 R_6 0 0.0 float16\n64 S_13 0 0.0 float16\n65 B_21 0 0.0 float16\n66 D_69 17657 1.91 float16\n67 B_22 43 0.0 float16\n68 D_70 9205 1.0 float16\n69 D_71 0 0.0 float16\n70 D_72 2064 0.22 float16\n71 S_15 0 0.0 float16\n72 B_23 0 0.0 float16\n73 D_73 910964 98.52 float16\n74 P_4 0 0.0 float16\n75 D_74 4179 0.45 float16\n76 D_75 0 0.0 float16\n77 D_76 831348 89.91 float16\n78 B_24 0 0.0 float16\n79 R_7 0 0.0 float16\n80 D_77 464536 50.24 float16\n81 B_25 1409 0.15 float16\n82 B_26 43 0.0 float16\n83 D_78 44612 4.82 float16\n84 D_79 4720 0.51 float16\n85 R_8 0 0.0 float16\n86 R_9 874673 94.6 float16\n87 S_16 0 0.0 float16\n88 D_80 4179 0.45 float16\n89 R_10 0 0.0 float16\n90 R_11 0 0.0 float16\n91 B_27 43 0.0 float16\n92 D_81 2087 0.23 float16\n93 D_82 689095 74.53 float16\n94 S_17 1409 0.15 float16\n95 R_12 0 0.0 float16\n96 B_28 0 0.0 float16\n97 R_13 0 0.0 float16\n98 D_83 17657 1.91 float16\n99 R_14 0 0.0 float16\n100 R_15 0 0.0 float16\n101 D_84 2209 0.24 float16\n102 R_16 0 0.0 float16\n103 B_29 436687 47.23 float16\n104 B_30 43 0.0 category\n105 S_18 0 0.0 float16\n106 D_86 122 0.01 float16\n107 D_87 923023 99.83 float16\n108 R_17 0 0.0 float16\n109 R_18 0 0.0 float16\n110 D_88 923072 99.83 float16\n111 B_31 0 0.0 float16\n112 S_19 0 0.0 float16\n113 R_19 0 0.0 float16\n114 B_32 0 0.0 float16\n115 S_20 0 0.0 float16\n116 R_20 0 0.0 float16\n117 R_21 0 0.0 float16\n118 B_33 43 0.0 float16\n119 D_89 2209 0.24 float16\n120 R_22 0 0.0 float16\n121 R_23 0 0.0 float16\n122 D_91 34513 3.73 float16\n123 D_92 0 0.0 float16\n124 D_93 0 0.0 float16\n125 D_94 0 0.0 float16\n126 R_24 0 0.0 float16\n127 R_25 0 0.0 float16\n128 D_96 0 0.0 float16\n129 S_22 4714 0.51 float16\n130 S_23 157 0.02 float16\n131 S_24 4634 0.5 float16\n132 S_25 2953 0.32 float16\n133 S_26 834 0.09 float16\n134 D_102 0 0.0 float16\n135 D_103 5050 0.55 float16\n136 D_104 5050 0.55 float16\n137 D_105 490747 53.08 float16\n138 D_106 803135 86.86 float16\n139 D_107 5050 0.55 float16\n140 B_36 0 0.0 float16\n141 B_37 834 0.09 float16\n142 R_26 775847 83.91 float16\n143 R_27 51391 5.56 float16\n144 B_38 43 0.0 category\n145 D_108 920135 99.51 float16\n146 D_109 877 0.09 float16\n147 D_110 914345 98.89 float16\n148 D_111 914345 98.89 float16\n149 B_39 912962 98.74 float16\n150 D_112 877 0.09 float16\n151 B_40 834 0.09 float16\n152 S_27 208690 22.57 float16\n153 D_113 17018 1.84 float16\n154 D_114 17018 1.84 category\n155 D_115 17018 1.84 float16\n156 D_116 17018 1.84 category\n157 D_117 17018 1.84 category\n158 D_118 17018 1.84 float16\n159 D_119 17018 1.84 float16\n160 D_120 17018 1.84 category\n161 D_121 17018 1.84 float16\n162 D_122 17018 1.84 float16\n163 D_123 17018 1.84 float16\n164 D_124 17018 1.84 float16\n165 D_125 17018 1.84 float16\n166 D_126 0 0.0 category\n167 D_127 0 0.0 float16\n168 D_128 5050 0.55 float16\n169 D_129 5050 0.55 float16\n170 B_41 834 0.09 float16\n171 B_42 906446 98.03 float16\n172 D_130 5050 0.55 float16\n173 D_131 5050 0.55 float16\n174 D_132 803056 86.85 float16\n175 D_133 0 0.0 float16\n176 R_28 0 0.0 float16\n177 D_134 897699 97.09 float16\n178 D_135 897699 97.09 float16\n179 D_136 897699 97.09 float16\n180 D_137 897699 97.09 float16\n181 D_138 897699 97.09 float16\n182 D_139 5050 0.55 float16\n183 D_140 0 0.0 float16\n184 D_141 5050 0.55 float16\n185 D_142 763770 82.6 float16\n186 D_143 5050 0.55 float16\n187 D_144 0 0.0 float16\n188 D_145 5050 0.55 float16","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
FeaturesNum of Missing valuesPercentageDataType
0S_200.0datetime64[ns]
1P_247840.52float16
2D_3900.0float16
3B_100.0float16
4B_2430.0float16
5R_100.0float16
6S_314428015.6float16
7D_41430.0float16
8B_3430.0float16
9D_4282757889.5float16
10D_4327221829.44float16
11D_44446124.82float16
12B_400.0float16
13D_45450.0float16
14B_500.0float16
15R_200.0float16
16D_4617926019.39float16
17D_4700.0float16
18D_4811698612.65float16
19D_4980263986.81float16
20B_61380.01float16
21B_700.0float16
22B_8123841.34float16
23D_5052986857.31float16
24D_5100.0float16
25B_900.0float16
26R_300.0float16
27D_5222090.24float16
28P_3304603.29float16
29B_1000.0float16
30D_5366102371.49float16
31S_500.0float16
32B_1100.0float16
33S_600.0float16
34D_54430.0float16
35R_400.0float16
36S_714428015.6float16
37B_1200.0float16
38S_800.0float16
39D_55604386.54float16
40D_5646322350.1float16
41B_1320100.22float16
42R_500.0float16
43D_5800.0float16
44S_920060821.7float16
45B_1400.0float16
46D_5989910.97float16
47D_6000.0float16
48D_6110158510.99float16
49B_1514090.15float16
50S_1100.0float16
51D_6215471316.73float16
52D_6300.0category
53D_6400.0category
54D_6500.0float16
55B_16430.0float16
56B_1748203652.13float16
57B_1800.0float16
58B_19430.0float16
59D_6681316887.95category
60B_20430.0float16
61D_68223802.42category
62S_126110.07float16
63R_600.0float16
64S_1300.0float16
65B_2100.0float16
66D_69176571.91float16
67B_22430.0float16
68D_7092051.0float16
69D_7100.0float16
70D_7220640.22float16
71S_1500.0float16
72B_2300.0float16
73D_7391096498.52float16
74P_400.0float16
75D_7441790.45float16
76D_7500.0float16
77D_7683134889.91float16
78B_2400.0float16
79R_700.0float16
80D_7746453650.24float16
81B_2514090.15float16
82B_26430.0float16
83D_78446124.82float16
84D_7947200.51float16
85R_800.0float16
86R_987467394.6float16
87S_1600.0float16
88D_8041790.45float16
89R_1000.0float16
90R_1100.0float16
91B_27430.0float16
92D_8120870.23float16
93D_8268909574.53float16
94S_1714090.15float16
95R_1200.0float16
96B_2800.0float16
97R_1300.0float16
98D_83176571.91float16
99R_1400.0float16
100R_1500.0float16
101D_8422090.24float16
102R_1600.0float16
103B_2943668747.23float16
104B_30430.0category
105S_1800.0float16
106D_861220.01float16
107D_8792302399.83float16
108R_1700.0float16
109R_1800.0float16
110D_8892307299.83float16
111B_3100.0float16
112S_1900.0float16
113R_1900.0float16
114B_3200.0float16
115S_2000.0float16
116R_2000.0float16
117R_2100.0float16
118B_33430.0float16
119D_8922090.24float16
120R_2200.0float16
121R_2300.0float16
122D_91345133.73float16
123D_9200.0float16
124D_9300.0float16
125D_9400.0float16
126R_2400.0float16
127R_2500.0float16
128D_9600.0float16
129S_2247140.51float16
130S_231570.02float16
131S_2446340.5float16
132S_2529530.32float16
133S_268340.09float16
134D_10200.0float16
135D_10350500.55float16
136D_10450500.55float16
137D_10549074753.08float16
138D_10680313586.86float16
139D_10750500.55float16
140B_3600.0float16
141B_378340.09float16
142R_2677584783.91float16
143R_27513915.56float16
144B_38430.0category
145D_10892013599.51float16
146D_1098770.09float16
147D_11091434598.89float16
148D_11191434598.89float16
149B_3991296298.74float16
150D_1128770.09float16
151B_408340.09float16
152S_2720869022.57float16
153D_113170181.84float16
154D_114170181.84category
155D_115170181.84float16
156D_116170181.84category
157D_117170181.84category
158D_118170181.84float16
159D_119170181.84float16
160D_120170181.84category
161D_121170181.84float16
162D_122170181.84float16
163D_123170181.84float16
164D_124170181.84float16
165D_125170181.84float16
166D_12600.0category
167D_12700.0float16
168D_12850500.55float16
169D_12950500.55float16
170B_418340.09float16
171B_4290644698.03float16
172D_13050500.55float16
173D_13150500.55float16
174D_13280305686.85float16
175D_13300.0float16
176R_2800.0float16
177D_13489769997.09float16
178D_13589769997.09float16
179D_13689769997.09float16
180D_13789769997.09float16
181D_13889769997.09float16
182D_13950500.55float16
183D_14000.0float16
184D_14150500.55float16
185D_14276377082.6float16
186D_14350500.55float16
187D_14400.0float16
188D_14550500.55float16
\n
"},"metadata":{}}]},{"cell_type":"markdown","source":"# Drop unuseful columns","metadata":{}},{"cell_type":"code","source":"test_dataset = test_dataset.drop(['S_2','D_42','D_49','D_66','D_73','D_76','R_9','B_29','D_87','D_88','D_106','R_26','D_108','D_110','D_111','B_39','B_42','D_132','D_134','D_135','D_136','D_137','D_138','D_142'], axis=1)","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:22.483216Z","iopub.execute_input":"2022-11-16T10:07:22.483785Z","iopub.status.idle":"2022-11-16T10:07:23.609363Z","shell.execute_reply.started":"2022-11-16T10:07:22.483736Z","shell.execute_reply":"2022-11-16T10:07:23.607874Z"},"trusted":true},"execution_count":47,"outputs":[]},{"cell_type":"markdown","source":"# Fill null values","metadata":{}},{"cell_type":"code","source":"selected_column = np.array(['P_2','S_3','B_2','D_41','D_43','B_3','D_44','D_45','D_46','D_48','D_50','D_53','S_7','D_56','S_9','S_12','S_17','B_6','B_8','D_52','P_3','D_54','D_55','B_13','D_59','D_61','B_15','D_62','B_16','B_17','D_77','B_19','B_20','D_69','B_22','D_70','D_72','D_74','R_7','B_25','B_26','D_78','D_79','D_80','B_27','D_81','R_12','D_82','D_105','S_27','D_83','R_14','D_84','D_86','R_20','B_33','D_89','D_91','S_22','S_23','S_24','S_25','S_26','D_102','D_103','D_104','D_107','B_37','R_27','D_109','D_112','B_40','D_113','D_115','D_118','D_119','D_121','D_122','D_123','D_124','D_125','D_128','D_129','B_41','D_130','D_131','D_133','D_139','D_140','D_141','D_143','D_144','D_145'])\n\nfor column in selected_column:\n test_dataset[column] = test_dataset[column].fillna(test_dataset[column].median())","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:23.611381Z","iopub.execute_input":"2022-11-16T10:07:23.611878Z","iopub.status.idle":"2022-11-16T10:07:27.634262Z","shell.execute_reply.started":"2022-11-16T10:07:23.611842Z","shell.execute_reply":"2022-11-16T10:07:27.631794Z"},"trusted":true},"execution_count":48,"outputs":[]},{"cell_type":"code","source":"selected_column2 = np.array(['D_68','B_30','B_38','D_114','D_116','D_117','D_120','D_126'])\n\nfor column2 in selected_column2:\n test_dataset[column2] = test_dataset[column2].fillna(test_dataset[column2].mode()[0])","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:27.636505Z","iopub.execute_input":"2022-11-16T10:07:27.637471Z","iopub.status.idle":"2022-11-16T10:07:27.754326Z","shell.execute_reply.started":"2022-11-16T10:07:27.637381Z","shell.execute_reply":"2022-11-16T10:07:27.752399Z"},"trusted":true},"execution_count":49,"outputs":[]},{"cell_type":"markdown","source":"# Check again null values","metadata":{}},{"cell_type":"code","source":"print(test_dataset.isnull().sum().to_string())","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:27.756869Z","iopub.execute_input":"2022-11-16T10:07:27.757294Z","iopub.status.idle":"2022-11-16T10:07:28.676026Z","shell.execute_reply.started":"2022-11-16T10:07:27.757258Z","shell.execute_reply":"2022-11-16T10:07:28.674604Z"},"trusted":true},"execution_count":50,"outputs":[{"name":"stdout","text":"P_2 0\nD_39 0\nB_1 0\nB_2 0\nR_1 0\nS_3 0\nD_41 0\nB_3 0\nD_43 0\nD_44 0\nB_4 0\nD_45 0\nB_5 0\nR_2 0\nD_46 0\nD_47 0\nD_48 0\nB_6 0\nB_7 0\nB_8 0\nD_50 0\nD_51 0\nB_9 0\nR_3 0\nD_52 0\nP_3 0\nB_10 0\nD_53 0\nS_5 0\nB_11 0\nS_6 0\nD_54 0\nR_4 0\nS_7 0\nB_12 0\nS_8 0\nD_55 0\nD_56 0\nB_13 0\nR_5 0\nD_58 0\nS_9 0\nB_14 0\nD_59 0\nD_60 0\nD_61 0\nB_15 0\nS_11 0\nD_62 0\nD_63 0\nD_64 0\nD_65 0\nB_16 0\nB_17 0\nB_18 0\nB_19 0\nB_20 0\nD_68 0\nS_12 0\nR_6 0\nS_13 0\nB_21 0\nD_69 0\nB_22 0\nD_70 0\nD_71 0\nD_72 0\nS_15 0\nB_23 0\nP_4 0\nD_74 0\nD_75 0\nB_24 0\nR_7 0\nD_77 0\nB_25 0\nB_26 0\nD_78 0\nD_79 0\nR_8 0\nS_16 0\nD_80 0\nR_10 0\nR_11 0\nB_27 0\nD_81 0\nD_82 0\nS_17 0\nR_12 0\nB_28 0\nR_13 0\nD_83 0\nR_14 0\nR_15 0\nD_84 0\nR_16 0\nB_30 0\nS_18 0\nD_86 0\nR_17 0\nR_18 0\nB_31 0\nS_19 0\nR_19 0\nB_32 0\nS_20 0\nR_20 0\nR_21 0\nB_33 0\nD_89 0\nR_22 0\nR_23 0\nD_91 0\nD_92 0\nD_93 0\nD_94 0\nR_24 0\nR_25 0\nD_96 0\nS_22 0\nS_23 0\nS_24 0\nS_25 0\nS_26 0\nD_102 0\nD_103 0\nD_104 0\nD_105 0\nD_107 0\nB_36 0\nB_37 0\nR_27 0\nB_38 0\nD_109 0\nD_112 0\nB_40 0\nS_27 0\nD_113 0\nD_114 0\nD_115 0\nD_116 0\nD_117 0\nD_118 0\nD_119 0\nD_120 0\nD_121 0\nD_122 0\nD_123 0\nD_124 0\nD_125 0\nD_126 0\nD_127 0\nD_128 0\nD_129 0\nB_41 0\nD_130 0\nD_131 0\nD_133 0\nR_28 0\nD_139 0\nD_140 0\nD_141 0\nD_143 0\nD_144 0\nD_145 0\n","output_type":"stream"}]},{"cell_type":"code","source":"test_dataset.shape","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:28.677945Z","iopub.execute_input":"2022-11-16T10:07:28.67881Z","iopub.status.idle":"2022-11-16T10:07:28.686928Z","shell.execute_reply.started":"2022-11-16T10:07:28.678759Z","shell.execute_reply":"2022-11-16T10:07:28.685767Z"},"trusted":true},"execution_count":51,"outputs":[{"execution_count":51,"output_type":"execute_result","data":{"text/plain":"(924621, 165)"},"metadata":{}}]},{"cell_type":"code","source":"test_dataset.head()","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:28.689192Z","iopub.execute_input":"2022-11-16T10:07:28.690304Z","iopub.status.idle":"2022-11-16T10:07:28.731517Z","shell.execute_reply.started":"2022-11-16T10:07:28.690242Z","shell.execute_reply":"2022-11-16T10:07:28.730132Z"},"trusted":true},"execution_count":52,"outputs":[{"execution_count":52,"output_type":"execute_result","data":{"text/plain":" P_2 D_39 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.568848 0.121399 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.841309 0.126465 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.697754 0.002724 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.513184 0.324707 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.254395 0.768066 \n\n B_1 B_2 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.010780 1.009766 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.016556 1.008789 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.001485 0.810059 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.149536 0.205688 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.563477 0.038025 \n\n R_1 S_3 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.006924 0.149414 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.009712 0.112183 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.002621 0.166138 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.002277 0.181152 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.502930 0.168335 \n\n D_41 B_3 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.000396 0.003576 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.006191 0.011383 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.004887 0.015945 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.005814 0.498535 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.009483 0.831055 \n\n D_43 D_44 ... \\\ncustomer_ID ... \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.007397 0.006786 ... \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.096863 0.004234 ... \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.105286 0.003382 ... \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.211670 0.258545 ... \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.071899 0.375732 ... \n\n D_130 D_131 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.004902 0.000975 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.001252 0.007633 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 1.003906 0.001152 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.009438 0.002775 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.003340 0.001000 \n\n D_133 R_28 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.006271 0.008820 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.002768 0.008789 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.002045 0.001852 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.009377 0.003622 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.007942 0.009232 \n\n D_139 D_140 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.005913 0.001250 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.004345 0.000866 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 1.000977 0.008896 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 1.007812 0.003754 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.006622 0.001140 \n\n D_141 D_143 \\\ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.006542 0.009163 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.009117 0.002197 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.895996 1.009766 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.919922 1.007812 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.009529 0.009407 \n\n D_144 D_145 \ncustomer_ID \n00000469ba478561f23a92a868bd366de6f6527a684c9a2... 0.003691 0.003220 \n00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397... 0.000247 0.007778 \n0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5... 0.457764 0.092041 \n00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf... 0.500977 0.182983 \n00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a... 0.001557 0.000525 \n\n[5 rows x 165 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
P_2D_39B_1B_2R_1S_3D_41B_3D_43D_44...D_130D_131D_133R_28D_139D_140D_141D_143D_144D_145
customer_ID
00000469ba478561f23a92a868bd366de6f6527a684c9a2e78fb826dcac3b9b70.5688480.1213990.0107801.0097660.0069240.1494140.0003960.0035760.0073970.006786...0.0049020.0009750.0062710.0088200.0059130.0012500.0065420.0091630.0036910.003220
00001bf2e77ff879fab36aa4fac689b9ba411dae63ae397d4263dafa1daedef50.8413090.1264650.0165561.0087890.0097120.1121830.0061910.0113830.0968630.004234...0.0012520.0076330.0027680.0087890.0043450.0008660.0091170.0021970.0002470.007778
0000210045da4f81e5f122c6bde5c2a617d03eef67f82c5e400fc98e7bd43ce80.6977540.0027240.0014850.8100590.0026210.1661380.0048870.0159450.1052860.003382...1.0039060.0011520.0020450.0018521.0009770.0088960.8959961.0097660.4577640.092041
00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976cf6e56734528702d6940.5131840.3247070.1495360.2056880.0022770.1811520.0058140.4985350.2116700.258545...0.0094380.0027750.0093770.0036221.0078120.0037540.9199221.0078120.5009770.182983
00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9a4693dd914fca225570.2543950.7680660.5634770.0380250.5029300.1683350.0094830.8310550.0718990.375732...0.0033400.0010000.0079420.0092320.0066220.0011400.0095290.0094070.0015570.000525
\n

5 rows × 165 columns

\n
"},"metadata":{}}]},{"cell_type":"markdown","source":"# Convert categorical variable to numbers","metadata":{}},{"cell_type":"code","source":"enc = OrdinalEncoder()\ncategorical_cols.remove('D_66')\n\ntrain_dataset[categorical_cols] = enc.fit_transform(train_dataset[categorical_cols])\ntest_dataset[categorical_cols] = enc.transform(test_dataset[categorical_cols])","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:28.734813Z","iopub.execute_input":"2022-11-16T10:07:28.735822Z","iopub.status.idle":"2022-11-16T10:07:31.055387Z","shell.execute_reply.started":"2022-11-16T10:07:28.735768Z","shell.execute_reply":"2022-11-16T10:07:31.053515Z"},"trusted":true},"execution_count":53,"outputs":[]},{"cell_type":"markdown","source":"# Scale the features by normalization","metadata":{}},{"cell_type":"code","source":"scaler = StandardScaler()\nX = pd.DataFrame(scaler.fit_transform(X), index=X.index, columns=X.columns) ","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Remove highly correlated features","metadata":{}},{"cell_type":"markdown","source":"Remove columns if there are > 90% of correlations","metadata":{}},{"cell_type":"code","source":"train_dataset_without_target = train_dataset.drop([\"target\"],axis=1)\n\ncor_matrix = train_dataset_without_target.corr()\ncol_core = set()\n\nfor i in range(len(cor_matrix.columns)):\n for j in range(i):\n if(cor_matrix.iloc[i, j] > 0.9):\n col_name = cor_matrix.columns[i]\n col_core.add(col_name)\ncol_core","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:07:31.057193Z","iopub.execute_input":"2022-11-16T10:07:31.057695Z","iopub.status.idle":"2022-11-16T10:08:07.264262Z","shell.execute_reply.started":"2022-11-16T10:07:31.057658Z","shell.execute_reply":"2022-11-16T10:08:07.262934Z"},"trusted":true},"execution_count":54,"outputs":[{"execution_count":54,"output_type":"execute_result","data":{"text/plain":"{'B_11',\n 'B_13',\n 'B_15',\n 'B_23',\n 'B_33',\n 'B_37',\n 'D_104',\n 'D_119',\n 'D_141',\n 'D_143',\n 'D_74',\n 'D_75',\n 'D_77',\n 'S_24',\n 'S_7'}"},"metadata":{}}]},{"cell_type":"code","source":"train_dataset = train_dataset.drop(col_core, axis=1)\ntest_dataset = test_dataset.drop(col_core, axis=1)","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:08:07.266419Z","iopub.execute_input":"2022-11-16T10:08:07.267956Z","iopub.status.idle":"2022-11-16T10:08:08.548849Z","shell.execute_reply.started":"2022-11-16T10:08:07.267896Z","shell.execute_reply":"2022-11-16T10:08:08.547577Z"},"trusted":true},"execution_count":55,"outputs":[]},{"cell_type":"code","source":"train_dataset.shape","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:08:08.550415Z","iopub.execute_input":"2022-11-16T10:08:08.550825Z","iopub.status.idle":"2022-11-16T10:08:08.559455Z","shell.execute_reply.started":"2022-11-16T10:08:08.550791Z","shell.execute_reply":"2022-11-16T10:08:08.557945Z"},"trusted":true},"execution_count":56,"outputs":[{"execution_count":56,"output_type":"execute_result","data":{"text/plain":"(458913, 151)"},"metadata":{}}]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Train Model","metadata":{}},{"cell_type":"code","source":"num_columns = [col for col in train_dataset.columns if col not in [\"target\"]]\n\nX = train_dataset[num_columns]\ny = train_dataset['target']\n\nprint(f\"X shape is = {X.shape}\" )\nprint(f\"Y shape is = {y.shape}\" )","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:08:08.561617Z","iopub.execute_input":"2022-11-16T10:08:08.56217Z","iopub.status.idle":"2022-11-16T10:08:08.998546Z","shell.execute_reply.started":"2022-11-16T10:08:08.56212Z","shell.execute_reply":"2022-11-16T10:08:08.996317Z"},"trusted":true},"execution_count":57,"outputs":[{"name":"stdout","text":"X shape is = (458913, 150)\nY shape is = (458913,)\n","output_type":"stream"}]},{"cell_type":"code","source":"x_train,x_test,y_train,y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)\n\nprint(f\"X_train shape is = {x_train.shape}\" )\nprint(f\"Y_train shape is = {y_train.shape}\" )\nprint(f\"X_test shape is = {x_test.shape}\" )\nprint(f\"Y_test shape is = {y_test.shape}\" )","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:08:09.000821Z","iopub.execute_input":"2022-11-16T10:08:09.001784Z","iopub.status.idle":"2022-11-16T10:08:10.555266Z","shell.execute_reply.started":"2022-11-16T10:08:09.001721Z","shell.execute_reply":"2022-11-16T10:08:10.553813Z"},"trusted":true},"execution_count":58,"outputs":[{"name":"stdout","text":"X_train shape is = (367130, 150)\nY_train shape is = (367130,)\nX_test shape is = (91783, 150)\nY_test shape is = (91783,)\n","output_type":"stream"}]},{"cell_type":"code","source":"\n\nd_train = lgb.Dataset(x_train, label=y_train, categorical_feature = categorical_cols)\n\nparams = {'objective': 'binary','n_estimators': 1200,'metric': 'binary_logloss','boosting': 'gbdt','num_leaves': 90,'reg_lambda' : 50,'colsample_bytree': 0.19,'learning_rate': 0.03,'min_child_samples': 2400,'max_bins': 511,'seed': 42,'verbose': -1}\n\n# trained model with 100 iterations\nmodel = lgb.train(params, d_train, 100)","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:08:10.561884Z","iopub.execute_input":"2022-11-16T10:08:10.562304Z","iopub.status.idle":"2022-11-16T10:10:19.765522Z","shell.execute_reply.started":"2022-11-16T10:08:10.562271Z","shell.execute_reply":"2022-11-16T10:10:19.763975Z"},"trusted":true},"execution_count":59,"outputs":[{"output_type":"display_data","data":{"text/plain":"","text/html":"\n"},"metadata":{}}]},{"cell_type":"markdown","source":"# Make Prediction","metadata":{}},{"cell_type":"code","source":"predictions = model.predict(test_dataset[num_columns])\npredictions","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:10:19.767656Z","iopub.execute_input":"2022-11-16T10:10:19.768148Z","iopub.status.idle":"2022-11-16T10:11:36.920637Z","shell.execute_reply.started":"2022-11-16T10:10:19.768095Z","shell.execute_reply":"2022-11-16T10:11:36.919364Z"},"trusted":true},"execution_count":60,"outputs":[{"execution_count":60,"output_type":"execute_result","data":{"text/plain":"array([0.01388529, 0.00215751, 0.03523218, ..., 0.54971303, 0.18039528,\n 0.04396771])"},"metadata":{}}]},{"cell_type":"markdown","source":"# Output","metadata":{}},{"cell_type":"code","source":"sample_dataset = pd.read_csv('/kaggle/input/amex-default-prediction/sample_submission.csv')\noutput = pd.DataFrame({'customer_ID': sample_dataset.customer_ID, 'prediction': predictions})\noutput.to_csv('submission.csv', index=False)","metadata":{"execution":{"iopub.status.busy":"2022-11-16T10:11:36.922259Z","iopub.execute_input":"2022-11-16T10:11:36.923261Z","iopub.status.idle":"2022-11-16T10:11:43.102332Z","shell.execute_reply.started":"2022-11-16T10:11:36.92322Z","shell.execute_reply":"2022-11-16T10:11:43.101051Z"},"trusted":true},"execution_count":61,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}