{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"#

๐Ÿฌ๐Ÿ”ง Data Pre-processing, EDA & Feature Engineering ๐Ÿ“‰
American Express

\n
","metadata":{}},{"cell_type":"markdown","source":"#
1. | Introduction ๐Ÿ‘‹
\n
\n \"Mart\"\n
\n
","metadata":{}},{"cell_type":"markdown","source":"##
Data Set Problems ๐Ÿค”
\n
\n ๐Ÿ‘‰ American Express is a globally integrated payments company. The largest payment card issuer in the world, they provide customers with access to products, insights, and experiences that enrich lives and build business success.
\n ๐Ÿ‘‰ Weโ€™ll be apply our machine learning skills to predict credit default which allows lenders to optimize lending decisions.
\n ๐Ÿ‘‰ Data pre-processing and feature engineering will be performed to prepare the dataset before it is used by the machine learning model.\n
\n\n##
Objectives of Notebook ๐Ÿ“Œ
\n
\n ๐Ÿ‘‰ This notebook aims to:\n \n
","metadata":{}},{"cell_type":"markdown","source":"##
Data Set Description ๐Ÿงพ
\n
\n ๐Ÿ‘‰ 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 \nwith the following features being categorical:\n\n['B_30', 'B_38', 'D_114', 'D_116', 'D_117', 'D_120', 'D_126', 'D_63', 'D_64', 'D_66', 'D_68'] \n\nOur task is to predict, for each customer_ID, the probability of a future payment default (target = 1).\n**Note that the negative class has been subsampled for this dataset at 5%, and thus receives a 20x weighting in the scoring metric.**\n ","metadata":{}},{"cell_type":"markdown","source":"#
2. | Importing Libraries ๐Ÿ“š
\n
\n ๐Ÿ‘‰ Importing libraries that will be used in this notebook.\n
","metadata":{}},{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport seaborn as sns\nimport matplotlib.pyplot as plt \nimport missingno as mso\nimport plotly.graph_objects as go\nimport plotly.offline as po\nfrom plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot\nimport plotly.express as px\nimport random\nimport plotly.figure_factory as ff\nfrom plotly.subplots import make_subplots\n","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-07-31T09:38:45.280979Z","iopub.execute_input":"2022-07-31T09:38:45.281837Z","iopub.status.idle":"2022-07-31T09:38:50.176927Z","shell.execute_reply.started":"2022-07-31T09:38:45.281731Z","shell.execute_reply":"2022-07-31T09:38:50.175952Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#
3. | Reading Dataset ๐Ÿ‘“
\n
\n ๐Ÿ‘‰ After importing libraries, the dataset that will be used will be imported.\n
","metadata":{}},{"cell_type":"code","source":"train_data=pd.read_feather(\"/kaggle/input/amexfeather/train_data.ftr\")\ntrain_labels=pd.read_csv(\"/kaggle/input/amex-default-prediction/train_labels.csv\")","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:38:56.027829Z","iopub.execute_input":"2022-07-31T09:38:56.028171Z","iopub.status.idle":"2022-07-31T09:39:16.761388Z","shell.execute_reply.started":"2022-07-31T09:38:56.028138Z","shell.execute_reply":"2022-07-31T09:39:16.760356Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Reading Dataset ---\ntrain_data.head().style.background_gradient(cmap='Greens').set_properties(**{'font-family': 'Segoe UI'}).hide_index()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:27.950187Z","iopub.execute_input":"2022-07-31T09:39:27.950556Z","iopub.status.idle":"2022-07-31T09:39:28.462299Z","shell.execute_reply.started":"2022-07-31T09:39:27.950523Z","shell.execute_reply":"2022-07-31T09:39:28.461404Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Reading Dataset ---\ntrain_labels.head().style.background_gradient(cmap='Greens').set_properties(**{'font-family': 'Segoe UI'}).hide_index()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:33.142019Z","iopub.execute_input":"2022-07-31T09:39:33.142853Z","iopub.status.idle":"2022-07-31T09:39:33.159724Z","shell.execute_reply.started":"2022-07-31T09:39:33.142814Z","shell.execute_reply":"2022-07-31T09:39:33.158735Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#
4. | Initial Data Exploration ๐Ÿ”
\n
\n ๐Ÿ‘‰ This section will focused on initial data exploration before pre-process the data.\n
","metadata":{}},{"cell_type":"markdown","source":"#
4.1 | Train Labels Data Exploration ๐Ÿ”
\n
\n ๐Ÿ‘‰ This section will focused on Train Labels data set exploration before pre-process the data.\n
","metadata":{}},{"cell_type":"code","source":"# Shape of the train label data set\n\nprint('\\033[1m'\"Shape of the train label data file\\n\"'\\033[0m',train_labels.shape)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:39.113094Z","iopub.execute_input":"2022-07-31T09:39:39.113505Z","iopub.status.idle":"2022-07-31T09:39:39.119324Z","shell.execute_reply.started":"2022-07-31T09:39:39.113469Z","shell.execute_reply":"2022-07-31T09:39:39.118228Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"## Data Type\nprint('\\033[1m'\"Data types of each column in train label data file\\n\"'\\033[0m',train_labels.dtypes)\n","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:43.258231Z","iopub.execute_input":"2022-07-31T09:39:43.258623Z","iopub.status.idle":"2022-07-31T09:39:43.265066Z","shell.execute_reply.started":"2022-07-31T09:39:43.258589Z","shell.execute_reply":"2022-07-31T09:39:43.264153Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#Missing value in the table \n\nprint('\\033[1m'\"Missing value present in each column of train label data file\\n\"'\\033[0m',train_labels.isna().any())\n","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:48.492115Z","iopub.execute_input":"2022-07-31T09:39:48.492482Z","iopub.status.idle":"2022-07-31T09:39:48.544515Z","shell.execute_reply.started":"2022-07-31T09:39:48.492445Z","shell.execute_reply":"2022-07-31T09:39:48.543536Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Check for the duplicated values \n\nprint('\\033[1m'\"Duplicate value present in each column of train label data file\\n\"'\\033[0m',train_labels.customer_ID.duplicated().any())\n\n","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:52.003866Z","iopub.execute_input":"2022-07-31T09:39:52.004213Z","iopub.status.idle":"2022-07-31T09:39:52.083771Z","shell.execute_reply.started":"2022-07-31T09:39:52.004182Z","shell.execute_reply":"2022-07-31T09:39:52.082698Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# No of unique customers \n\nprint('\\033[1m'\"No of Unique customers in the train label data file\\n\"'\\033[0m',train_labels.customer_ID.nunique())\n","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:55.261062Z","iopub.execute_input":"2022-07-31T09:39:55.261432Z","iopub.status.idle":"2022-07-31T09:39:55.452931Z","shell.execute_reply.started":"2022-07-31T09:39:55.261382Z","shell.execute_reply":"2022-07-31T09:39:55.451864Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Count of the Target\n\nprint('\\033[1m'\"Value count of the target column in the train label data file\\n\"'\\033[0m',train_labels.target.value_counts())\n","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:39:58.989604Z","iopub.execute_input":"2022-07-31T09:39:58.990096Z","iopub.status.idle":"2022-07-31T09:39:59.002747Z","shell.execute_reply.started":"2022-07-31T09:39:58.990051Z","shell.execute_reply":"2022-07-31T09:39:59.001496Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"target=train_labels.target.value_counts(normalize=True)\ntarget.rename(index={1:'Default',0:'Paid'},inplace=True)\ncolors = ['#17becf', '#E1396C']\ndata = go.Pie(\nvalues= target,\nlabels= target.index,\nmarker=dict(colors=colors),\ntextinfo='label+percent'\n)\nlayout = go.Layout(\ntitle=dict(text = \"Target Distribution\",x=0.46,y=0.95,font_size=20)\n)\nfig = go.Figure(data=data,layout=layout)\nfig.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:02.055632Z","iopub.execute_input":"2022-07-31T09:40:02.056050Z","iopub.status.idle":"2022-07-31T09:40:02.225956Z","shell.execute_reply.started":"2022-07-31T09:40:02.056013Z","shell.execute_reply":"2022-07-31T09:40:02.221499Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del train_labels","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:08.513727Z","iopub.execute_input":"2022-07-31T09:40:08.514063Z","iopub.status.idle":"2022-07-31T09:40:08.519047Z","shell.execute_reply.started":"2022-07-31T09:40:08.514032Z","shell.execute_reply":"2022-07-31T09:40:08.518035Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#
4.1 | Train Data Exploration ๐Ÿ”
\n
\n ๐Ÿ‘‰ This section will focused on Train data set exploration before pre-process the data.\n
","metadata":{}},{"cell_type":"code","source":"# Shape of the train data set\n\nprint('\\033[1m'\"Shape of the train data file\\n\"'\\033[0m',train_data.shape)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:20.569403Z","iopub.execute_input":"2022-07-31T09:40:20.570292Z","iopub.status.idle":"2022-07-31T09:40:20.576487Z","shell.execute_reply.started":"2022-07-31T09:40:20.570240Z","shell.execute_reply":"2022-07-31T09:40:20.575164Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"## Data Type\nprint('\\033[1m'\"Data types of each column in train label data file\\n\"'\\033[0m',train_data.dtypes)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:23.755973Z","iopub.execute_input":"2022-07-31T09:40:23.756326Z","iopub.status.idle":"2022-07-31T09:40:23.766542Z","shell.execute_reply.started":"2022-07-31T09:40:23.756293Z","shell.execute_reply":"2022-07-31T09:40:23.765562Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_data.info(max_cols=200, show_counts=True)\n","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:27.450067Z","iopub.execute_input":"2022-07-31T09:40:27.450432Z","iopub.status.idle":"2022-07-31T09:40:32.829598Z","shell.execute_reply.started":"2022-07-31T09:40:27.450383Z","shell.execute_reply":"2022-07-31T09:40:32.828643Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Following are the categorical variables 'B_30', 'B_38', 'D_114', 'D_116', 'D_117', 'D_120', 'D_126', 'D_63', 'D_64', 'D_66', 'D_68'. We are going to analyze it w.r.t to Target","metadata":{}},{"cell_type":"markdown","source":"#
4.1.1 | Analyzing the Balance categorical variables ๐Ÿ”
\n
\n ๐Ÿ‘‰ This section will focused on Analyzing the Balance categorical variables before pre-process the data.\n
\n \n ","metadata":{}},{"cell_type":"code","source":"train_data[\"B_30\"].value_counts()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:41.784361Z","iopub.execute_input":"2022-07-31T09:40:41.785078Z","iopub.status.idle":"2022-07-31T09:40:41.831295Z","shell.execute_reply.started":"2022-07-31T09:40:41.785039Z","shell.execute_reply":"2022-07-31T09:40:41.830378Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\nblack_grad = ['#100C07', '#3E3B39', '#6D6A6A', '#9B9A9C', '#CAC9CD']\ncyan_grad = ['#142459', '#176BA0', '#19AADE', '#1AC9E6', '#87EAFA']\ncolors=cyan_grad\nlabels=train_data['B_30'].dropna().unique()\norder=train_data['B_30'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(16, 8))\nplt.suptitle('B_30 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='B_30', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+100,rect.get_height(), horizontalalignment='center',\n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\n\nplt.xlabel('B_30 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['B_30'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%',\n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre)\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 29)\nprint('\\033[1m'+'.: B_30 Content Total :.'+'\\033[0m')\nprint('\\033[36m*' * 29+'\\033[0m')\ntrain_data.B_38.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:46.032744Z","iopub.execute_input":"2022-07-31T09:40:46.033090Z","iopub.status.idle":"2022-07-31T09:40:46.678828Z","shell.execute_reply.started":"2022-07-31T09:40:46.033059Z","shell.execute_reply":"2022-07-31T09:40:46.677775Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\nblack_grad = ['#100C07', '#3E3B39', '#6D6A6A', '#9B9A9C', '#CAC9CD']\ncyan_grad = ['#142459', '#176BA0', '#19AADE', '#1AC9E6', '#87EAFA']\ncolors=cyan_grad\nlabels=train_data['B_38'].dropna().unique()\norder=train_data['B_38'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(16, 8))\nplt.suptitle('B_38 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='B_38', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+100,rect.get_height(), horizontalalignment='center',\n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\n\nplt.xlabel('B_38 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['B_38'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%',\n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre)\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 29)\nprint('\\033[1m'+'.: B_38 Content Total :.'+'\\033[0m')\nprint('\\033[36m*' * 29+'\\033[0m')\ntrain_data.B_38.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:40:52.428224Z","iopub.execute_input":"2022-07-31T09:40:52.428904Z","iopub.status.idle":"2022-07-31T09:40:53.092221Z","shell.execute_reply.started":"2022-07-31T09:40:52.428863Z","shell.execute_reply":"2022-07-31T09:40:53.091340Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#
4.1.2| Analyzing the Delinquency categorical variables ๐Ÿ”
\n
\n ๐Ÿ‘‰ This section will focused on Analyzing the Delinquency categorical variables before pre-process the data.\n
\n\n\n\n","metadata":{}},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_114'].dropna().unique()\norder=train_data['D_114'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_114 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_114', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_114 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_114'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_114 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_114.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:05.460181Z","iopub.execute_input":"2022-07-31T09:41:05.460552Z","iopub.status.idle":"2022-07-31T09:41:06.066691Z","shell.execute_reply.started":"2022-07-31T09:41:05.460519Z","shell.execute_reply":"2022-07-31T09:41:06.065471Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_116'].dropna().unique()\norder=train_data['D_116'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_116 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_116', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_116 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_116'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_116 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_116.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:13.138930Z","iopub.execute_input":"2022-07-31T09:41:13.139753Z","iopub.status.idle":"2022-07-31T09:41:14.569160Z","shell.execute_reply.started":"2022-07-31T09:41:13.139711Z","shell.execute_reply":"2022-07-31T09:41:14.568250Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_117'].dropna().unique()\norder=train_data['D_117'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_117 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_117', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_117 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_117'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_117 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_117.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:21.257334Z","iopub.execute_input":"2022-07-31T09:41:21.258424Z","iopub.status.idle":"2022-07-31T09:41:21.937718Z","shell.execute_reply.started":"2022-07-31T09:41:21.258375Z","shell.execute_reply":"2022-07-31T09:41:21.936594Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_120'].dropna().unique()\norder=train_data['D_120'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_120 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_120', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_120 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_120'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_120 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_120.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:30.677951Z","iopub.execute_input":"2022-07-31T09:41:30.678348Z","iopub.status.idle":"2022-07-31T09:41:31.273308Z","shell.execute_reply.started":"2022-07-31T09:41:30.678316Z","shell.execute_reply":"2022-07-31T09:41:31.272338Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_126'].dropna().unique()\norder=train_data['D_126'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_126 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_126', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_126 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_126'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_126 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_126.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:38.300204Z","iopub.execute_input":"2022-07-31T09:41:38.300574Z","iopub.status.idle":"2022-07-31T09:41:38.942549Z","shell.execute_reply.started":"2022-07-31T09:41:38.300541Z","shell.execute_reply":"2022-07-31T09:41:38.941468Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_63'].dropna().unique()\norder=train_data['D_63'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_63 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_63', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_63 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_63'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_63 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_63.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:45.053936Z","iopub.execute_input":"2022-07-31T09:41:45.054291Z","iopub.status.idle":"2022-07-31T09:41:45.740331Z","shell.execute_reply.started":"2022-07-31T09:41:45.054258Z","shell.execute_reply":"2022-07-31T09:41:45.739340Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_64'].dropna().unique()\norder=train_data['D_64'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_64 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_64', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_64 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_64'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_64 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_64.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:51.182341Z","iopub.execute_input":"2022-07-31T09:41:51.183028Z","iopub.status.idle":"2022-07-31T09:41:51.837719Z","shell.execute_reply.started":"2022-07-31T09:41:51.182989Z","shell.execute_reply":"2022-07-31T09:41:51.836709Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_66'].dropna().unique()\norder=train_data['D_66'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_66 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_66', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_66 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_66'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_66 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_66.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:41:58.348369Z","iopub.execute_input":"2022-07-31T09:41:58.349062Z","iopub.status.idle":"2022-07-31T09:41:58.814292Z","shell.execute_reply.started":"2022-07-31T09:41:58.349021Z","shell.execute_reply":"2022-07-31T09:41:58.813379Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Setting Colors, Labels, Order ---\npurple_grad = ['#491D8B', '#6929C4', '#8A3FFC', '#A56EFF', '#BE95FF']\ncolors=purple_grad\nlabels=train_data['D_68'].dropna().unique()\norder=train_data['D_68'].value_counts().index\n\n# --- Size for Both Figures ---\nplt.figure(figsize=(18, 8))\nplt.suptitle('D_68 Distribution', fontweight='heavy', fontsize='16', fontfamily='sans-serif', \n color=black_grad[0])\n\n# --- Histogram ---\ncountplt = plt.subplot(1, 2, 1)\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_68', data=train_data, palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nfor rect in ax.patches:\n ax.text (rect.get_x()+rect.get_width()/2, rect.get_height()+20,rect.get_height(), horizontalalignment='center', \n fontsize=12, bbox=dict(facecolor='none', edgecolor=black_grad[0], linewidth=0.15, boxstyle='round'))\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_68 Distribution', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\ncountplt\n\n# --- Pie Chart ---\nplt.subplot(1, 2, 2)\nplt.title('Pie Chart', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nplt.pie(train_data['D_68'].value_counts(), colors=colors, labels=order, pctdistance=0.67, autopct='%.2f%%', \n wedgeprops=dict(alpha=0.8, edgecolor=black_grad[1]), textprops={'fontsize':12})\ncentre=plt.Circle((0, 0), 0.45, fc='white', edgecolor=black_grad[1])\nplt.gcf().gca().add_artist(centre);\n\n# --- Count Categorical Labels w/out Dropping Null Walues ---\nprint('\\033[36m*' * 30)\nprint('\\033[1m'+'.: D_68 Total :.'+'\\033[0m')\nprint('\\033[36m*' * 30+'\\033[0m')\ntrain_data.D_68.value_counts(dropna=False)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:05.079935Z","iopub.execute_input":"2022-07-31T09:42:05.080273Z","iopub.status.idle":"2022-07-31T09:42:05.773296Z","shell.execute_reply.started":"2022-07-31T09:42:05.080243Z","shell.execute_reply":"2022-07-31T09:42:05.772392Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#
4.1.3| Analyzing the categorical variables w.r.t Target ๐Ÿ”
\n
\n ๐Ÿ‘‰ This section will focused on Analyzing the categorical variables w.r.t Target before pre-process the data.\n
\n","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='B_30', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('B_30 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:12.888329Z","iopub.execute_input":"2022-07-31T09:42:12.889365Z","iopub.status.idle":"2022-07-31T09:42:13.599290Z","shell.execute_reply.started":"2022-07-31T09:42:12.889320Z","shell.execute_reply":"2022-07-31T09:42:13.598390Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='B_38', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('B_38 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:19.248249Z","iopub.execute_input":"2022-07-31T09:42:19.248751Z","iopub.status.idle":"2022-07-31T09:42:20.332246Z","shell.execute_reply.started":"2022-07-31T09:42:19.248706Z","shell.execute_reply":"2022-07-31T09:42:20.331325Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_114', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_114 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:24.875087Z","iopub.execute_input":"2022-07-31T09:42:24.875451Z","iopub.status.idle":"2022-07-31T09:42:25.539350Z","shell.execute_reply.started":"2022-07-31T09:42:24.875398Z","shell.execute_reply":"2022-07-31T09:42:25.538334Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_116', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_116 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:30.116800Z","iopub.execute_input":"2022-07-31T09:42:30.117148Z","iopub.status.idle":"2022-07-31T09:42:30.910512Z","shell.execute_reply.started":"2022-07-31T09:42:30.117115Z","shell.execute_reply":"2022-07-31T09:42:30.909460Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_117', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_117 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:35.905654Z","iopub.execute_input":"2022-07-31T09:42:35.906557Z","iopub.status.idle":"2022-07-31T09:42:36.553395Z","shell.execute_reply.started":"2022-07-31T09:42:35.906517Z","shell.execute_reply":"2022-07-31T09:42:36.552371Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_120', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_120 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:42.099499Z","iopub.execute_input":"2022-07-31T09:42:42.099977Z","iopub.status.idle":"2022-07-31T09:42:42.838308Z","shell.execute_reply.started":"2022-07-31T09:42:42.099935Z","shell.execute_reply":"2022-07-31T09:42:42.836493Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_126', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_126 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:49.528980Z","iopub.execute_input":"2022-07-31T09:42:49.529338Z","iopub.status.idle":"2022-07-31T09:42:50.201638Z","shell.execute_reply.started":"2022-07-31T09:42:49.529304Z","shell.execute_reply":"2022-07-31T09:42:50.200718Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_66', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_66 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:55.045661Z","iopub.execute_input":"2022-07-31T09:42:55.046040Z","iopub.status.idle":"2022-07-31T09:42:55.516290Z","shell.execute_reply.started":"2022-07-31T09:42:55.046003Z","shell.execute_reply":"2022-07-31T09:42:55.515294Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\nplt.title('Histogram', fontweight='bold', fontsize=14, fontfamily='sans-serif', color=black_grad[0])\nax = sns.countplot(x='D_68', data=train_data,hue='target', palette=colors, order=order, edgecolor=black_grad[2], alpha=0.85)\nplt.tight_layout(rect=[0, 0.04, 1, 0.965])\nplt.xlabel('D_68 Distribution w.r.t target', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.ylabel('Total', fontweight='bold', fontsize=11, fontfamily='sans-serif', color=black_grad[1])\nplt.grid(axis='y', alpha=0.4)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:42:59.980991Z","iopub.execute_input":"2022-07-31T09:42:59.981332Z","iopub.status.idle":"2022-07-31T09:43:00.693324Z","shell.execute_reply.started":"2022-07-31T09:42:59.981301Z","shell.execute_reply":"2022-07-31T09:43:00.692334Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#
4.1.4| Analyzing the numerical attributes ๐Ÿ”
\n
\n ๐Ÿ‘‰ This section will focused on Analyzing the numerical attributes before pre-process the data.\n
\n","metadata":{}},{"cell_type":"markdown","source":"## Data analysis on the Balance attributes","metadata":{}},{"cell_type":"code","source":"# Categorcal columns \n\ncategorical_cols=['B_30', 'B_38', 'D_63', 'D_64', 'D_66', 'D_68',\n 'D_114', 'D_116', 'D_117', 'D_120', 'D_126', 'target']","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:43:14.422199Z","iopub.execute_input":"2022-07-31T09:43:14.422567Z","iopub.status.idle":"2022-07-31T09:43:14.427432Z","shell.execute_reply.started":"2022-07-31T09:43:14.422535Z","shell.execute_reply":"2022-07-31T09:43:14.426471Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"\ncols=[col for col in train_data.columns if (col.startswith(('B','T'))) & (col not in categorical_cols[:-1])]\n\ntrain_balance_cols =train_data[cols]","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:43:24.600495Z","iopub.execute_input":"2022-07-31T09:43:24.600847Z","iopub.status.idle":"2022-07-31T09:43:25.266796Z","shell.execute_reply.started":"2022-07-31T09:43:24.600816Z","shell.execute_reply":"2022-07-31T09:43:25.265797Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_balance_cols.select_dtypes(exclude='object').describe().T.style.background_gradient(cmap='RdPu').set_properties(**{'font-family': 'Segoe UI'})","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:43:30.524518Z","iopub.execute_input":"2022-07-31T09:43:30.525088Z","iopub.status.idle":"2022-07-31T09:44:01.515850Z","shell.execute_reply.started":"2022-07-31T09:43:30.525049Z","shell.execute_reply":"2022-07-31T09:44:01.514940Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Plot Missing Values ---\nmso.bar(train_balance_cols, fontsize=9, color=[purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0],\n purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[1], purple_grad[1]], \n figsize=(15, 8), sort='descending', labels=True)\n\n# --- Title & Subtitle Settings ---\nplt.suptitle('Missing Values in each Columns', fontweight='heavy', x=0.124, y=1.22, ha='left',fontsize='16', \n fontfamily='sans-serif', color=black_grad[0])\nplt.title('Almost all columns have missing value.\\n\\nThe total of missing values in each column is less than 25%, which means that imputation can still be done to fill in the missing values in the\\ntwo columns.', \n fontsize='8', fontfamily='sans-serif', loc='left', color=black_grad[1], pad=5)\nplt.grid(axis='both', alpha=0);\n\n# --- Total Missing Values in each Columns ---\nprint('\\033[36m*' * 43)\nprint('\\033[1m'+'.: Total Missing Values in each Columns :.'+'\\033[0m')\nprint('\\033[36m*' * 43+'\\033[0m')\ntrain_balance_cols.isnull().sum()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:44:15.387818Z","iopub.execute_input":"2022-07-31T09:44:15.388166Z","iopub.status.idle":"2022-07-31T09:44:21.794383Z","shell.execute_reply.started":"2022-07-31T09:44:15.388134Z","shell.execute_reply":"2022-07-31T09:44:21.793361Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#plot distribution for selected B variables with respect to target label\nnrows = 8\nncols = 5\nfig, axes = plt.subplots(figsize=(24,22)) \nb_columns=['B_1', 'B_2', 'B_3', 'B_4', 'B_5', 'B_6', 'B_7', 'B_8', 'B_9', 'B_10',\n 'B_11', 'B_12', 'B_13', 'B_14', 'B_15', 'B_16', 'B_17', 'B_18', 'B_19',\n 'B_20', 'B_21', 'B_22', 'B_23', 'B_24', 'B_25', 'B_26', 'B_27', 'B_28',\n 'B_29', 'B_31', 'B_32', 'B_33', 'B_36', 'B_37', 'B_39', 'B_40', 'B_41',\n 'B_42']\nfor i, col in enumerate(b_columns):\n ax=fig.add_subplot(nrows, ncols, i+1)\n sns.kdeplot(x=train_data[col],hue=train_data['target'], multiple=\"stack\",palette=[\"#FF3333\" ,\"#00CC00\"],ax=ax)\n \nfig.tight_layout() \nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:45:32.202917Z","iopub.execute_input":"2022-07-31T09:45:32.203260Z","iopub.status.idle":"2022-07-31T09:57:01.261070Z","shell.execute_reply.started":"2022-07-31T09:45:32.203229Z","shell.execute_reply":"2022-07-31T09:57:01.260238Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del train_data[col]","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:57:09.661201Z","iopub.execute_input":"2022-07-31T09:57:09.662162Z","iopub.status.idle":"2022-07-31T09:57:09.668574Z","shell.execute_reply.started":"2022-07-31T09:57:09.662110Z","shell.execute_reply":"2022-07-31T09:57:09.667543Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"corr=train_balance_cols.corr()\nplt.figure(figsize=(24,22))\nax = sns.heatmap(corr,cmap=\"YlGnBu\", linewidths=.5, vmin=-1, vmax=1, center=0, annot=True, annot_kws={'fontsize':12,'fontweight':'bold'}, cbar=False,fmt='.2f' )\nplt.yticks(rotation=0)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:44:39.347562Z","iopub.execute_input":"2022-07-31T09:44:39.347903Z","iopub.status.idle":"2022-07-31T09:45:03.630565Z","shell.execute_reply.started":"2022-07-31T09:44:39.347873Z","shell.execute_reply":"2022-07-31T09:45:03.629510Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del train_balance_cols","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Data analysis on the Spend attributes","metadata":{}},{"cell_type":"code","source":"cols=[col for col in train_data.columns if (col.startswith(('S','T'))) & (col not in categorical_cols[:-1])]\n\ntrain_spend_cols =train_data[cols]","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:05:20.223494Z","iopub.execute_input":"2022-07-30T07:05:20.223859Z","iopub.status.idle":"2022-07-30T07:05:20.627219Z","shell.execute_reply.started":"2022-07-30T07:05:20.223824Z","shell.execute_reply":"2022-07-30T07:05:20.626198Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_spend_cols.select_dtypes(exclude='object').describe().T.style.background_gradient(cmap='RdPu').set_properties(**{'font-family': 'Segoe UI'})","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:05:20.628862Z","iopub.execute_input":"2022-07-30T07:05:20.629249Z","iopub.status.idle":"2022-07-30T07:05:38.428201Z","shell.execute_reply.started":"2022-07-30T07:05:20.629211Z","shell.execute_reply":"2022-07-30T07:05:38.426867Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Plot Missing Values ---\nmso.bar(train_spend_cols, fontsize=9, color=[purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0],\n purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[1], purple_grad[1]], \n figsize=(15, 8), sort='descending', labels=True)\n\n# --- Title & Subtitle Settings ---\nplt.suptitle('Missing Values in each Columns', fontweight='heavy', x=0.124, y=1.22, ha='left',fontsize='16', \n fontfamily='sans-serif', color=black_grad[0])\nplt.title('Almost all columns have no missing value except \\n\\nThe total of missing values in each column is less than 25%, which means that imputation can still be done to fill in the missing values in the\\nthe columns.', \n fontsize='8', fontfamily='sans-serif', loc='left', color=black_grad[1], pad=5)\nplt.grid(axis='both', alpha=0);\n\n# --- Total Missing Values in each Columns ---\nprint('\\033[36m*' * 43)\nprint('\\033[1m'+'.: Total Missing Values in each Columns :.'+'\\033[0m')\nprint('\\033[36m*' * 43+'\\033[0m')\ntrain_spend_cols.isnull().sum()","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:05:38.429990Z","iopub.execute_input":"2022-07-30T07:05:38.430407Z","iopub.status.idle":"2022-07-30T07:05:42.383793Z","shell.execute_reply.started":"2022-07-30T07:05:38.430368Z","shell.execute_reply":"2022-07-30T07:05:42.382670Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#plot distribution for selected B variables with respect to target label\nnrows = 5\nncols = 5\nfig, axes = plt.subplots(figsize=(24,22)) \ns_columns=['S_2', 'S_3', 'S_5', 'S_6', 'S_7', 'S_8', 'S_9', 'S_11', 'S_12', 'S_13',\n 'S_15', 'S_16', 'S_17', 'S_18', 'S_19', 'S_20', 'S_22', 'S_23', 'S_24',\n 'S_25', 'S_26', 'S_27']\nfor i, col in enumerate(s_columns):\n ax=fig.add_subplot(nrows, ncols, i+1)\n sns.kdeplot(x=train_data[col],hue=train_data['target'], multiple=\"stack\",palette=[\"#FF3333\" ,\"#00CC00\"],ax=ax)\n \nfig.tight_layout() \nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:05:42.394794Z","iopub.execute_input":"2022-07-30T07:05:42.395225Z","iopub.status.idle":"2022-07-30T07:12:50.443086Z","shell.execute_reply.started":"2022-07-30T07:05:42.395187Z","shell.execute_reply":"2022-07-30T07:12:50.441858Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del train_data[col]","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"corr=train_spend_cols.corr()\nplt.figure(figsize=(24,22))\nax = sns.heatmap(corr,cmap=\"YlGnBu\", linewidths=.5, vmin=-1, vmax=1, center=0, annot=True, annot_kws={'fontsize':12,'fontweight':'bold'}, cbar=False,fmt='.2f' )\nplt.yticks(rotation=0)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:12:50.445053Z","iopub.execute_input":"2022-07-30T07:12:50.445418Z","iopub.status.idle":"2022-07-30T07:12:59.745560Z","shell.execute_reply.started":"2022-07-30T07:12:50.445359Z","shell.execute_reply":"2022-07-30T07:12:59.744627Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del train_spend_cols","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Data analysis on the Deliquency attributes","metadata":{}},{"cell_type":"code","source":"cols=[col for col in train_data.columns if (col.startswith(('D','T'))) & (col not in categorical_cols[:-1])]\n\ntrain_deliquency_cols =train_data[cols]","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:12:59.746961Z","iopub.execute_input":"2022-07-30T07:12:59.747476Z","iopub.status.idle":"2022-07-30T07:13:01.315813Z","shell.execute_reply.started":"2022-07-30T07:12:59.747442Z","shell.execute_reply":"2022-07-30T07:13:01.314793Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_deliquency_cols.select_dtypes(exclude='object').describe().T.style.background_gradient(cmap='RdPu').set_properties(**{'font-family': 'Segoe UI'})","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:13:01.317279Z","iopub.execute_input":"2022-07-30T07:13:01.317637Z","iopub.status.idle":"2022-07-30T07:14:14.280534Z","shell.execute_reply.started":"2022-07-30T07:13:01.317600Z","shell.execute_reply":"2022-07-30T07:14:14.279276Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# --- Plot Missing Values ---\nmso.bar(train_deliquency_cols, fontsize=9, color=[purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0],\n purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[0], purple_grad[1], purple_grad[1]], \n figsize=(24, 22), sort='descending', labels=True)\n\n# --- Title & Subtitle Settings ---\nplt.suptitle('Missing Values in each Columns', fontweight='heavy', x=0.124, y=1.22, ha='left',fontsize='16', \n fontfamily='sans-serif', color=black_grad[0])\nplt.title('Almost all columns have no missing value except \\n\\nThe total of missing values in each column is less than 25%, which means that imputation can still be done to fill in the missing values in the\\nthe columns.', \n fontsize='8', fontfamily='sans-serif', loc='left', color=black_grad[1], pad=5)\nplt.grid(axis='both', alpha=0);\n\n# --- Total Missing Values in each Columns ---\nprint('\\033[36m*' * 43)\nprint('\\033[1m'+'.: Total Missing Values in each Columns :.'+'\\033[0m')\nprint('\\033[36m*' * 43+'\\033[0m')\ntrain_deliquency_cols.isnull().sum()","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:14:14.281956Z","iopub.execute_input":"2022-07-30T07:14:14.282408Z","iopub.status.idle":"2022-07-30T07:14:31.995502Z","shell.execute_reply.started":"2022-07-30T07:14:14.282372Z","shell.execute_reply":"2022-07-30T07:14:31.994605Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#plot distribution for selected B variables with respect to target label\nnrows = 18\nncols = 5\nfig, axes = plt.subplots(figsize=(24,22)) \nd_columns=['D_39', 'D_41', 'D_42', 'D_43', 'D_44', 'D_45', 'D_46', 'D_47', 'D_48',\n 'D_49', 'D_50', 'D_51', 'D_52', 'D_53', 'D_54', 'D_55', 'D_56', 'D_58',\n 'D_59', 'D_60', 'D_61', 'D_62', 'D_65', 'D_69', 'D_70', 'D_71', 'D_72',\n 'D_73', 'D_74', 'D_75', 'D_76', 'D_77', 'D_78', 'D_79', 'D_80', 'D_81',\n 'D_82', 'D_83', 'D_84', 'D_86', 'D_88', 'D_89', 'D_91', 'D_92',\n 'D_93', 'D_94', 'D_96', 'D_102', 'D_103', 'D_104', 'D_105',\n 'D_107', 'D_108', 'D_109', 'D_110', 'D_111', 'D_112', 'D_113', 'D_115',\n 'D_118', 'D_119', 'D_121', 'D_122', 'D_123', 'D_124', 'D_125', 'D_127',\n 'D_128', 'D_129', 'D_130', 'D_131', 'D_132', 'D_133', 'D_134', 'D_135',\n 'D_136', 'D_137', 'D_138', 'D_139', 'D_140', 'D_141', 'D_142', 'D_143',\n 'D_144', 'D_145']\nfor i, col in enumerate(d_columns):\n ax=fig.add_subplot(nrows, ncols, i+1)\n sns.kdeplot(x=train_data[col],hue=train_data['target'], multiple=\"stack\",palette=[\"#FF3333\" ,\"#00CC00\"],ax=ax)\n \nfig.tight_layout() \nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-30T07:14:32.007849Z","iopub.execute_input":"2022-07-30T07:14:32.008483Z","iopub.status.idle":"2022-07-30T07:37:39.247282Z","shell.execute_reply.started":"2022-07-30T07:14:32.008448Z","shell.execute_reply":"2022-07-30T07:37:39.246425Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del train_data[col]","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del train_deliquency_cols","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Work In progress","metadata":{}},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}