{"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":"# H&M Recommendation - Majority Voting Ensemble\n\n\n\n\nThank you for your checking this notebook.\n\nThis is my notebook for \"H&M Personalized Fashion Recommendations\" competition [(Link)](https://www.kaggle.com/c/h-and-m-personalized-fashion-recommendations/overview).\n\nThe end of competition is approaching now and would like to share one way of ensemble \"Majority Voting\" to improve our score as much as possible in the last minutes.\n\nIf you think this notebook is interesting for you, please leave your comment or question and I appreciate your upvote as well. :) \n\n\n## Contents\n1. [Import Library & Set Config](#config)\n2. [Load Data](#load)\n3. [Prepare Data & Submission](#prepare)\n6. [Conclution](#conclution)\n7. [Reference](#ref)","metadata":{},"attachments":{"10d0f9eb-f600-417a-8cfa-d71566deee8e.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"\n\n---\n## 1. Import Library & Set Config\n---\n\n[Back to Contents](#top)","metadata":{}},{"cell_type":"code","source":"# === General ===\nimport sys, warnings, time, os, copy, gc, re, random, pickle, cudf\nwarnings.filterwarnings('ignore')\nfrom IPython.display import display\nimport matplotlib.pyplot as plt\nimport numpy as np\nimport pandas as pd\n# pd.set_option('display.max_rows', 50)\n# pd.set_option('display.max_columns', None)\n# pd.set_option(\"display.max_colwidth\", 10000)\nimport seaborn as sns\nsns.set()\nfrom pandas.io.json import json_normalize\nfrom pprint import pprint\nfrom pathlib import Path\nfrom tqdm import tqdm\ntqdm.pandas()\nfrom collections import Counter","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:24:27.226796Z","iopub.execute_input":"2022-04-17T04:24:27.227344Z","iopub.status.idle":"2022-04-17T04:24:27.236538Z","shell.execute_reply.started":"2022-04-17T04:24:27.227309Z","shell.execute_reply":"2022-04-17T04:24:27.235732Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"DEBUG = False","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:24:27.940276Z","iopub.execute_input":"2022-04-17T04:24:27.941970Z","iopub.status.idle":"2022-04-17T04:24:27.949716Z","shell.execute_reply.started":"2022-04-17T04:24:27.941925Z","shell.execute_reply":"2022-04-17T04:24:27.948866Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n\n---\n## 2. Load Data\n\nIn this notebook, I try to ensemble 5 submittion files from notebooks below. \nOne is mine and others are published. Please check and upvote them too.\n\n1. [H&M EDA & Rule Base by Customer Age](https://www.kaggle.com/code/hechtjp/h-m-eda-rule-base-by-customer-age)\n2. [H&M : Framework for Partitioned Validation](https://www.kaggle.com/code/negoto/h-m-framework-for-partitioned-validation)\n3. [Trending](https://www.kaggle.com/code/ebn7amdi/trending)\n4. [[LB 0.0236] Ensemble gives you Bronze medal](https://www.kaggle.com/code/jaloeffe92/lb-0-0236-ensemble-gives-you-bronze-medal)\n5. [H&M Ensembling - with LSTM be7226](https://www.kaggle.com/code/qianyetang/h-m-ensembling-with-lstm-be7226)\n\n---\n\n[Back to Contents](#top)","metadata":{}},{"cell_type":"code","source":"def display_df(df, head=3):\n print(f'The shape of df is {df.shape}.\\n')\n display(df.head(head))","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:30:41.530847Z","iopub.execute_input":"2022-04-17T04:30:41.531106Z","iopub.status.idle":"2022-04-17T04:30:41.535404Z","shell.execute_reply.started":"2022-04-17T04:30:41.531078Z","shell.execute_reply":"2022-04-17T04:30:41.534353Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSub1 = cudf.read_csv('../input/h-m-eda-rule-base-by-customer-age/submission.csv')\ndisplay_df(dfSub1, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:31:00.283297Z","iopub.execute_input":"2022-04-17T04:31:00.283555Z","iopub.status.idle":"2022-04-17T04:31:05.145987Z","shell.execute_reply.started":"2022-04-17T04:31:00.283528Z","shell.execute_reply":"2022-04-17T04:31:05.145285Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSub2 = cudf.read_csv('../input/h-m-framework-for-partitioned-validation/submission.csv')\ndisplay_df(dfSub2, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:31:14.493479Z","iopub.execute_input":"2022-04-17T04:31:14.493739Z","iopub.status.idle":"2022-04-17T04:31:18.819164Z","shell.execute_reply.started":"2022-04-17T04:31:14.493709Z","shell.execute_reply":"2022-04-17T04:31:18.818542Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSub3 = cudf.read_csv('../input/trending/submission.csv')\ndisplay_df(dfSub3, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:31:32.479316Z","iopub.execute_input":"2022-04-17T04:31:32.479565Z","iopub.status.idle":"2022-04-17T04:31:36.346419Z","shell.execute_reply.started":"2022-04-17T04:31:32.479537Z","shell.execute_reply":"2022-04-17T04:31:36.345658Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSub4 = cudf.read_csv('../input/lb-0-0236-ensemble-gives-you-bronze-medal/submission.csv')\ndisplay_df(dfSub4, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:32:46.056564Z","iopub.execute_input":"2022-04-17T04:32:46.056852Z","iopub.status.idle":"2022-04-17T04:32:46.285884Z","shell.execute_reply.started":"2022-04-17T04:32:46.056801Z","shell.execute_reply":"2022-04-17T04:32:46.285165Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSub5 = cudf.read_csv('../input/h-m-ensembling-with-lstm-be7226/submission.csv')\ndisplay_df(dfSub5, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:32:47.315446Z","iopub.execute_input":"2022-04-17T04:32:47.315727Z","iopub.status.idle":"2022-04-17T04:32:47.538118Z","shell.execute_reply.started":"2022-04-17T04:32:47.315695Z","shell.execute_reply":"2022-04-17T04:32:47.537328Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSub1.columns = ['customer_id', 'prediction1']\ndfSub2.columns = ['customer_id', 'prediction2']\ndfSub3.columns = ['customer_id', 'prediction3']\ndfSub4.columns = ['customer_id', 'prediction4']\ndfSub5.columns = ['customer_id', 'prediction5']","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:33:14.882067Z","iopub.execute_input":"2022-04-17T04:33:14.882336Z","iopub.status.idle":"2022-04-17T04:33:14.888288Z","shell.execute_reply.started":"2022-04-17T04:33:14.882307Z","shell.execute_reply":"2022-04-17T04:33:14.887626Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n\n---\n## 3. Check top 12 articles whih are the most voted & create submission file.\n- Using \"Counter\" to count how many times each articles appeared and select top 12 ones.\n\n---\n\n[Back to Contents](#top)","metadata":{}},{"cell_type":"code","source":"dfSub = dfSub1.merge(dfSub2, on='customer_id', how='left')\ndfSub = dfSub.merge(dfSub3, on='customer_id', how='left')\ndfSub = dfSub.merge(dfSub4, on='customer_id', how='left')\ndfSub = dfSub.merge(dfSub5, on='customer_id', how='left')\ndisplay_df(dfSub, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:36:56.429506Z","iopub.execute_input":"2022-04-17T04:36:56.430285Z","iopub.status.idle":"2022-04-17T04:36:56.771430Z","shell.execute_reply.started":"2022-04-17T04:36:56.430244Z","shell.execute_reply":"2022-04-17T04:36:56.770728Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"if DEBUG:\n dfSub = dfSub.sample(frac=0.001, random_state=7)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:37:13.555269Z","iopub.execute_input":"2022-04-17T04:37:13.555525Z","iopub.status.idle":"2022-04-17T04:37:13.559733Z","shell.execute_reply.started":"2022-04-17T04:37:13.555495Z","shell.execute_reply":"2022-04-17T04:37:13.558877Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def select_topn(x, n=12):\n listX = str(x).split()\n c = Counter(listX)\n values, counts = zip(*c.most_common(n))\n listY = ' '.join(values)\n return listY ","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:37:15.797097Z","iopub.execute_input":"2022-04-17T04:37:15.797723Z","iopub.status.idle":"2022-04-17T04:37:15.802568Z","shell.execute_reply.started":"2022-04-17T04:37:15.797687Z","shell.execute_reply":"2022-04-17T04:37:15.801770Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSub['pred_sum'] = dfSub['prediction1'] + ' ' + dfSub['prediction2'] + ' ' + dfSub['prediction3'] + ' ' + dfSub['prediction4'] + ' ' + dfSub['prediction5']\n\ndfSub = dfSub.to_pandas()\n\ndfSub['pred_top12'] = dfSub['pred_sum'].progress_apply(select_topn)\n\ndisplay_df(dfSub, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:37:58.658785Z","iopub.execute_input":"2022-04-17T04:37:58.659319Z","iopub.status.idle":"2022-04-17T04:38:48.626138Z","shell.execute_reply.started":"2022-04-17T04:37:58.659285Z","shell.execute_reply":"2022-04-17T04:38:48.625415Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(dfSub['prediction1'][248891])\nprint('')\nprint(dfSub['prediction2'][248891])\nprint('')\nprint(dfSub['prediction3'][248891])\nprint('')\nprint(dfSub['prediction4'][248891])\nprint('')\nprint(dfSub['prediction5'][248891])\nprint('')\nprint(dfSub['pred_sum'][248891])\nprint('')\nprint(dfSub['pred_top12'][248891])","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:38:48.627923Z","iopub.execute_input":"2022-04-17T04:38:48.628417Z","iopub.status.idle":"2022-04-17T04:38:48.639258Z","shell.execute_reply.started":"2022-04-17T04:38:48.628381Z","shell.execute_reply":"2022-04-17T04:38:48.638567Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSampleSub = dfSub[['customer_id', 'pred_top12']]\ndisplay_df(dfSampleSub, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:38:48.640503Z","iopub.execute_input":"2022-04-17T04:38:48.641008Z","iopub.status.idle":"2022-04-17T04:38:49.392395Z","shell.execute_reply.started":"2022-04-17T04:38:48.640971Z","shell.execute_reply":"2022-04-17T04:38:49.390897Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfSampleSub.columns = ['customer_id', 'prediction']\n\ndfSampleSub.to_csv(f'submission.csv', index=False)\nprint(f'Saved submission.csv.')","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:38:49.394348Z","iopub.execute_input":"2022-04-17T04:38:49.394614Z","iopub.status.idle":"2022-04-17T04:39:00.673579Z","shell.execute_reply.started":"2022-04-17T04:38:49.394577Z","shell.execute_reply":"2022-04-17T04:39:00.672800Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfCheck = cudf.read_csv('./submission.csv')\ndisplay_df(dfCheck, head=3)","metadata":{"execution":{"iopub.status.busy":"2022-04-17T04:39:00.675275Z","iopub.execute_input":"2022-04-17T04:39:00.675700Z","iopub.status.idle":"2022-04-17T04:39:00.917388Z","shell.execute_reply.started":"2022-04-17T04:39:00.675651Z","shell.execute_reply":"2022-04-17T04:39:00.913867Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n\n---\n\n## 4. Conclution\n\nThis is the most simple way of ensemble and there are a lot of other way to do that.\n\nIf you have any proposal or advice for me, please feel free to leave the comment!\n\nThank you for your reading through this Notebook!\n\nIf you think this notebook is interesting for you, please do click upvote :)\n\n---\n\n[Back to Contents](#top)","metadata":{}},{"cell_type":"markdown","source":"\n\n---\n## 5. Reference\n\n1. [H&M: Faster Trending Products Weekly](https://www.kaggle.com/hervind/h-m-faster-trending-products-weekly)\n2. [Trending](https://www.kaggle.com/hechtjp/trending?scriptVersionId=89904417)\n\n---\n\n[Back to Contents](#top)","metadata":{}},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}