{"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] Postal-code Grouping\n\n**This is the sequel version to my previous notebook below.**  \nhttps://www.kaggle.com/code/junjitakeshima/h-m-easy-grouping-by-sex-attribute-age-en-jp  \n\n**[Original Aim]**  \nThe target period for this competition's scoring is 1week of late September. This \"late September\" is a bit subtle timing, because Autmn is coming in some countries(e.g. Sweden where H&M HQ exists) while other countries are still enjoying Summer (e.g. Indonesia). If such Season can be grouped, the score could be improved.   \n\nスコアの対象となるのは9月下旬の1週間に売れた商品。この9月下旬というのが微妙で、秋が始まりつつある国もあれば（例えばH&Mの本国スウェーデン）、常夏の国もあるはず（例えばインドネシア）。そうした季節感の異なる地域毎にグルーピングをすればスコアは改善するのではないか。  \n\n![image.png](attachment:74e98156-13a7-4959-a951-612190002a99.png)\n\n**[Approach]**  \nFirst Approach  : tried to find \"Country\" information in postal_code (... fail)  \n　　　　　　　　　　　(but, I found the first 10 digits of 64 can make postal_code unique)  \n           \nSecond Approach :  \n(1) Define the postal code area (maybe \"town\") where \"Sweater\" or \"Coat\" are sold in September as winter-coming area  \n(2) Grouping by customers in winter-coming area and others  \n(3) Recommend most-sold items grouped by Age, Sex/attribute, and Area.  \n\nThe result was not as expected (slightly got worse from 0.0078->0.0069).\n\n最初のアプローチ：　郵便番号の中に国情報が入っているのではないかと考え、それを見つけようとした（が、断念。。）  \n　　　　　　　　　　（ただ、郵便番号64桁のうち最初の10桁でユニークになっていることに気づき、その10桁をエリアコードとして使いました）  \n次のアプローチ　：  \n(1) 9月に入ってからセーターまたはコートが売れている郵便番号エリアを冬間近地域として定義  \n(2) 顧客を冬間近地域に住む顧客とそれ以外に分類  \n(3) 年齢・性別/属性・地域区別に応じてグルーピングして最も売れたアイテムをおススメ\n\n結果としては当初の狙い通りのスコアにはならなかった（0.0078→0.0069へと僅かに悪化）","metadata":{},"attachments":{"74e98156-13a7-4959-a951-612190002a99.png":{"image/png":"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"}}},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns; sns.set()\nfrom datetime import datetime, date, timedelta\n\nfrom collections import Counter, defaultdict\nfrom PIL import Image\nfrom pathlib import Path\n\npath = Path(\"/kaggle/input/h-and-m-personalized-fashion-recommendations/\")\npd.set_option(\"display.max_columns\", None)","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-03-22T05:09:16.65374Z","iopub.execute_input":"2022-03-22T05:09:16.654514Z","iopub.status.idle":"2022-03-22T05:09:17.800385Z","shell.execute_reply.started":"2022-03-22T05:09:16.654387Z","shell.execute_reply":"2022-03-22T05:09:17.79934Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 1. Reading Dataframes","metadata":{}},{"cell_type":"code","source":"transactions_df = pd.read_csv(path / \"transactions_train.csv\", dtype = {'article_id': str})\narticles_df = pd.read_csv(path / \"articles.csv\", dtype = {'article_id': str})\ncustomers_df = pd.read_csv(path / \"customers.csv\")\nsubmission = pd.read_csv(path / \"sample_submission.csv\")","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:09:17.802178Z","iopub.execute_input":"2022-03-22T05:09:17.802436Z","iopub.status.idle":"2022-03-22T05:10:46.97073Z","shell.execute_reply.started":"2022-03-22T05:09:17.802403Z","shell.execute_reply":"2022-03-22T05:10:46.969581Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"transactions_df[\"t_dat\"] = pd.to_datetime(transactions_df['t_dat'])\ntransactions_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:21:57.440219Z","iopub.execute_input":"2022-03-22T05:21:57.440557Z","iopub.status.idle":"2022-03-22T05:22:05.146614Z","shell.execute_reply.started":"2022-03-22T05:21:57.440517Z","shell.execute_reply":"2022-03-22T05:22:05.145592Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:10:46.972136Z","iopub.execute_input":"2022-03-22T05:10:46.972381Z","iopub.status.idle":"2022-03-22T05:10:47.011235Z","shell.execute_reply.started":"2022-03-22T05:10:46.972351Z","shell.execute_reply":"2022-03-22T05:10:47.010561Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 2. Insight for Postal Code\n\nAs you can see, the postal code is Hexadecimal(16) x 64digit. Why this long??  \nHowever, unique nymber is 352,899 while total count is 1,371,980. Let's see more in detail.  \n\nご覧の通り、郵便番号は16進数の64桁。なんでこんなに長いんだ？  \nただ、総数1,371,980に対し、ユニーク数は352,899。64桁全てを使ってユニークになっているわけではなさそう。もう少し見てみよう。","metadata":{}},{"cell_type":"code","source":"display(customers_df[\"postal_code\"].describe())\nprint(\"postal code length = \", len(customers_df.loc[0, \"postal_code\"]))","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:10:47.012865Z","iopub.execute_input":"2022-03-22T05:10:47.013205Z","iopub.status.idle":"2022-03-22T05:10:48.150283Z","shell.execute_reply.started":"2022-03-22T05:10:47.013176Z","shell.execute_reply":"2022-03-22T05:10:48.14928Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Let's split the postal code, and see how many digits make the postal code unique. \n\n64桁を一度バラして、何桁でユニークになっているのかを見てみる。","metadata":{}},{"cell_type":"code","source":"post_df = customers_df[\"postal_code\"].str.split('', expand=True)\npost_df[\"1_2\"] = post_df[1].str.cat(post_df[2])\nfor i in range(8) :\n    post_df[f\"1_{i+3}\"] = post_df[f\"1_{i+2}\"].str.cat(post_df[i+3])","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:14:17.065197Z","iopub.execute_input":"2022-03-22T05:14:17.066017Z","iopub.status.idle":"2022-03-22T05:14:58.982613Z","shell.execute_reply.started":"2022-03-22T05:14:17.065971Z","shell.execute_reply":"2022-03-22T05:14:58.981219Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Let's see the conbination of 1st to 2nd digit, 1st to 3rd・・・until 1st to 10th digit.  \nAs you can see, first 10digits make the postal code unique.  \n\n1-2桁目、1-3桁目・・という具合に1-10桁目までの組み合わせを見てみる。  \nご覧の通り、最初の10桁でユニークになっていることが分かる。","metadata":{}},{"cell_type":"code","source":"a = post_df[1].nunique()\nprint(\"1st dig: counts =\",  a)\nfor i in range(9) :\n    b = post_df[f\"1_{i+2}\"].nunique()\n    print(f\"1 to {i+2} : counts =\",  b, \"   above row *\", b/a)\n    a = b","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:15:15.323096Z","iopub.execute_input":"2022-03-22T05:15:15.323445Z","iopub.status.idle":"2022-03-22T05:15:18.499253Z","shell.execute_reply.started":"2022-03-22T05:15:15.323405Z","shell.execute_reply":"2022-03-22T05:15:18.498271Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 3. Try to classify customers who live in Winter-coming area and others\n\nBut failed to find the information to indicate/suggest \"country\", and eventually gave up..  \nContinue to try 2nd approach (try to define the area where winter items are sold as \"winter-coming area\").  \n\nただ、結局国を表す情報がどこにあるのかは見つからず、最初のアプローチ(国単位でグルーピング)は断念。  \n2番目のアプローチ（冬物が売れている地域を冬間近地域として定義）を試してみる。","metadata":{}},{"cell_type":"code","source":"articles_df.head()","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:15:55.23895Z","iopub.execute_input":"2022-03-22T05:15:55.23924Z","iopub.status.idle":"2022-03-22T05:15:55.263539Z","shell.execute_reply.started":"2022-03-22T05:15:55.239211Z","shell.execute_reply":"2022-03-22T05:15:55.262866Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"There are dozens of product-type names. Out of them, let's choose \"Sweater\" and \"Coat\" as typical winter-items.  \n製品タイプは何十個とあるが、この中から冬物として「セーター」と「コート」を使うことにする。","metadata":{}},{"cell_type":"code","source":"articles_df[\"product_type_name\"].unique()","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:18:13.919184Z","iopub.execute_input":"2022-03-22T05:18:13.920748Z","iopub.status.idle":"2022-03-22T05:18:13.941833Z","shell.execute_reply.started":"2022-03-22T05:18:13.92068Z","shell.execute_reply":"2022-03-22T05:18:13.941108Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Sweater or Coat items are more than 9000. I never thought such many items exist.  \nセーターまたはコートだけで9000アイテム以上。こんなにあるんだ。","metadata":{}},{"cell_type":"code","source":"articles_df[articles_df[\"product_type_name\"]==(\"Sweater\" or \"Coat\")]","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:18:33.555506Z","iopub.execute_input":"2022-03-22T05:18:33.556464Z","iopub.status.idle":"2022-03-22T05:18:33.636855Z","shell.execute_reply.started":"2022-03-22T05:18:33.556399Z","shell.execute_reply":"2022-03-22T05:18:33.635749Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df = pd.concat([customers_df, post_df[\"1_10\"].reset_index(drop = True)], axis=1)\ncustomers_df = customers_df.rename(columns = {\"1_10\" : \"town\"})\ncustomers_df = customers_df.drop(\"postal_code\", axis = 1)\ncustomers_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:19:06.579799Z","iopub.execute_input":"2022-03-22T05:19:06.580215Z","iopub.status.idle":"2022-03-22T05:19:08.15496Z","shell.execute_reply.started":"2022-03-22T05:19:06.580137Z","shell.execute_reply":"2022-03-22T05:19:08.153866Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"articles_df.loc[(articles_df[\"product_type_name\"]==(\"Sweater\" or \"Coat\")), \"season\"] = 1\narticles_df.fillna(0, inplace = True)","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:19:20.662978Z","iopub.execute_input":"2022-03-22T05:19:20.663856Z","iopub.status.idle":"2022-03-22T05:19:20.864046Z","shell.execute_reply.started":"2022-03-22T05:19:20.663782Z","shell.execute_reply":"2022-03-22T05:19:20.863335Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"custom_temp = pd.DataFrame(customers_df[[\"customer_id\", \"town\"]])\ntrans_temp  = transactions_df.copy()\ntrans_temp = pd.merge(trans_temp, custom_temp, on = \"customer_id\", how = \"left\")\n\narticle_temp = pd.DataFrame(articles_df[[\"article_id\", \"season\"]])\ntrans_temp = pd.merge(trans_temp, article_temp, on = \"article_id\", how = \"left\")\ntrans_temp = trans_temp.loc[trans_temp.t_dat >= pd.to_datetime('2020-09-01')]\n\nregion_df = trans_temp[[\"town\", \"season\"]].groupby(\"town\").sum().reset_index()","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:22:05.148655Z","iopub.execute_input":"2022-03-22T05:22:05.149718Z","iopub.status.idle":"2022-03-22T05:22:58.488839Z","shell.execute_reply.started":"2022-03-22T05:22:05.149667Z","shell.execute_reply":"2022-03-22T05:22:58.487811Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"it seems that the customers who live in the area where sweater/coat were sold in September-2020 are apprx 30% of the total customers.  \n9月に入ってセーター・コートが売れた地域に住んでいる顧客は顧客全体の3割程度のようだ。","metadata":{}},{"cell_type":"code","source":"region_df.loc[region_df[\"season\"]>10, \"season\"] = 10\nregion_df[\"season\"].plot.hist(bins=10)","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:22:58.490649Z","iopub.execute_input":"2022-03-22T05:22:58.490937Z","iopub.status.idle":"2022-03-22T05:22:58.876855Z","shell.execute_reply.started":"2022-03-22T05:22:58.490901Z","shell.execute_reply":"2022-03-22T05:22:58.875778Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"region_df.loc[(region_df[\"season\"] >0), \"season\"] = 1\n\nprint(\"number of regions which have transactions after 2020-09-01 = : \", len(region_df))\nprint(\"number of Winter-coming regions\", region_df[\"season\"].sum())","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:22:58.87858Z","iopub.execute_input":"2022-03-22T05:22:58.878821Z","iopub.status.idle":"2022-03-22T05:22:58.891215Z","shell.execute_reply.started":"2022-03-22T05:22:58.87879Z","shell.execute_reply":"2022-03-22T05:22:58.890284Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df = pd.merge(customers_df, region_df, on = \"town\", how =\"left\") \ncustomers_df[\"season\"].fillna(0, inplace = True)\n\ntemp = customers_df[\"season\"].value_counts().index.to_list()\nplt.figure(figsize=(5, 5))\nplt.rcParams[\"font.size\"] = 12\nplt.pie(customers_df[\"season\"].value_counts().sort_values(ascending=False), \n        labels = temp, startangle = 90, counterclock=False, autopct=\"%1.1f%%\")\nplt.show","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:22:58.893404Z","iopub.execute_input":"2022-03-22T05:22:58.893806Z","iopub.status.idle":"2022-03-22T05:23:00.540864Z","shell.execute_reply.started":"2022-03-22T05:22:58.893767Z","shell.execute_reply":"2022-03-22T05:23:00.539836Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 4. Grouping by Age\n\nThe contents hereafter is basically same as my previous notebook below, and hence I'll omit the explanation.\n\nここからは、前回のNotebookと基本的に同一ですので、解説は省略します。\n\nhttps://www.kaggle.com/code/junjitakeshima/h-m-easy-grouping-by-sex-attribute-age-en-jp","metadata":{}},{"cell_type":"code","source":"customers_df[\"age\"].plot.hist(bins=50)","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:23:00.542535Z","iopub.execute_input":"2022-03-22T05:23:00.543906Z","iopub.status.idle":"2022-03-22T05:23:01.106979Z","shell.execute_reply.started":"2022-03-22T05:23:00.543851Z","shell.execute_reply":"2022-03-22T05:23:01.105691Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"age_id = 0\nage_group = pd.DataFrame(columns =[\"age\",\"age_id\"])\nage=16\n\nfor i in range(53) :\n    if age < 30 :\n        temp_group = pd.DataFrame({\"age\":[age, age+1], \"age_id\":[age_id, age_id]})\n        age_group = age_group.append(temp_group)\n        age += 2\n        age_id += 1\n    elif  age < 60 :\n        temp_group = pd.DataFrame({\"age\":[age, age+1, age+2, age+3, age+4],\"age_id\":[age_id, age_id, age_id, age_id, age_id]})\n        age_group = age_group.append(temp_group)\n        age += 5\n        age_id += 1\n    else:\n        temp_group = pd.DataFrame({\"age\":[age], \"age_id\":[age_id]})\n        age_group = age_group.append(temp_group)\n        age += 1","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:23:01.108702Z","iopub.execute_input":"2022-03-22T05:23:01.109061Z","iopub.status.idle":"2022-03-22T05:23:01.173113Z","shell.execute_reply.started":"2022-03-22T05:23:01.109015Z","shell.execute_reply":"2022-03-22T05:23:01.172099Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"16-99 age will be grouped into age_id 0-13.   \n\n16~99歳の年齢を0～13までのage_idにグルーピング","metadata":{}},{"cell_type":"code","source":"age_group","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:23:01.174571Z","iopub.execute_input":"2022-03-22T05:23:01.174794Z","iopub.status.idle":"2022-03-22T05:23:01.18802Z","shell.execute_reply.started":"2022-03-22T05:23:01.174766Z","shell.execute_reply":"2022-03-22T05:23:01.186926Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df = pd.merge(customers_df, age_group, on=\"age\", how = \"left\")\ncustomers_df = customers_df.drop([\"FN\", \"Active\", \"club_member_status\", \"fashion_news_frequency\"], axis=1)\ncustomers_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:23:21.139335Z","iopub.execute_input":"2022-03-22T05:23:21.139705Z","iopub.status.idle":"2022-03-22T05:23:23.443827Z","shell.execute_reply.started":"2022-03-22T05:23:21.139657Z","shell.execute_reply":"2022-03-22T05:23:23.442873Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 5. Grouping by Sex/Attribute\n\n**Approach**  \n(1) Based on \"index_group_name\" in article list, Purchase history will be classified as \"Women's item\", \"Men's item\", \"Kid's item\", etc.  \n(2) Based on the purchase history, classify the customers into 5 attributes, guessing those who purchased more Women's items are Women, and those who purchased more Men's items are Men, etc.  \n\n**アプローチ**  \n(1) 品目リストのindex_group_nameをもとに、購入履歴を女性用品・男性用品・子供用品等に分類  \n(2) 女性用の購入履歴が多い顧客を女性、男性用の購入履歴が多い顧客を男性、という具合に顧客層を5つの属性に分類  \n　　（女性　／　ヤング(Divided)　／　男性　／　子持ち　／　スポーツパーソン）","metadata":{}},{"cell_type":"code","source":"articles_df.head(5)","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:23:50.859033Z","iopub.execute_input":"2022-03-22T05:23:50.859315Z","iopub.status.idle":"2022-03-22T05:23:50.886827Z","shell.execute_reply.started":"2022-03-22T05:23:50.859284Z","shell.execute_reply":"2022-03-22T05:23:50.885722Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(articles_df[\"index_group_name\"].unique())\nprint(articles_df[\"index_group_no\"].unique())","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:24:01.283026Z","iopub.execute_input":"2022-03-22T05:24:01.283725Z","iopub.status.idle":"2022-03-22T05:24:01.300617Z","shell.execute_reply.started":"2022-03-22T05:24:01.283685Z","shell.execute_reply":"2022-03-22T05:24:01.299333Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sex_category = articles_df[[\"index_group_no\", \"index_group_name\"]].reset_index()\ndisplay(sex_category[\"index_group_name\"].value_counts())","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:24:09.879257Z","iopub.execute_input":"2022-03-22T05:24:09.880325Z","iopub.status.idle":"2022-03-22T05:24:09.911261Z","shell.execute_reply.started":"2022-03-22T05:24:09.880244Z","shell.execute_reply":"2022-03-22T05:24:09.910327Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sex_category_list = sex_category[\"index_group_name\"].value_counts().index.to_list()\nplt.figure(figsize=(5, 5))\nplt.rcParams[\"font.size\"] = 12\nplt.pie(sex_category[\"index_group_name\"].value_counts().sort_values(ascending=False), \n        labels = sex_category_list, startangle = 90, counterclock=False, autopct=\"%1.1f%%\")\nplt.show","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:24:19.303778Z","iopub.execute_input":"2022-03-22T05:24:19.304355Z","iopub.status.idle":"2022-03-22T05:24:19.476288Z","shell.execute_reply.started":"2022-03-22T05:24:19.304301Z","shell.execute_reply":"2022-03-22T05:24:19.475174Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del sex_category_list","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:24:28.228915Z","iopub.execute_input":"2022-03-22T05:24:28.229203Z","iopub.status.idle":"2022-03-22T05:24:28.234466Z","shell.execute_reply.started":"2022-03-22T05:24:28.229174Z","shell.execute_reply":"2022-03-22T05:24:28.233134Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"articles_category_df = pd.DataFrame(articles_df[[\"article_id\", \"index_group_no\"]])\narticles_category_df.columns = [\"article_id\", \"sex_attribute\"]\narticles_category_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:24:36.007695Z","iopub.execute_input":"2022-03-22T05:24:36.008514Z","iopub.status.idle":"2022-03-22T05:24:36.027311Z","shell.execute_reply.started":"2022-03-22T05:24:36.008433Z","shell.execute_reply":"2022-03-22T05:24:36.026215Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"transactions_df = pd.merge(transactions_df, articles_category_df, on = \"article_id\", how = \"left\")\ntransactions_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:24:43.959005Z","iopub.execute_input":"2022-03-22T05:24:43.959423Z","iopub.status.idle":"2022-03-22T05:24:58.74859Z","shell.execute_reply.started":"2022-03-22T05:24:43.959393Z","shell.execute_reply":"2022-03-22T05:24:58.747686Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cust_sex = transactions_df[[\"customer_id\", \"sex_attribute\", \"article_id\"]].groupby([\"customer_id\",\"sex_attribute\"]).count().unstack()\ncust_sex.columns = [\"Woman\", \"Young\", \"Man\", \"Have-kids\", \"Sports-person\"]\ncust_sex","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:24:58.75065Z","iopub.execute_input":"2022-03-22T05:24:58.751171Z","iopub.status.idle":"2022-03-22T05:25:26.72137Z","shell.execute_reply.started":"2022-03-22T05:24:58.751124Z","shell.execute_reply":"2022-03-22T05:25:26.720636Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%%time\ncust_sex[\"attribute\"] = cust_sex.apply(lambda x : list(x[x == x.max()].index), axis=1)\ncust_sex","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:25:26.722892Z","iopub.execute_input":"2022-03-22T05:25:26.723224Z","iopub.status.idle":"2022-03-22T05:31:12.1541Z","shell.execute_reply.started":"2022-03-22T05:25:26.723194Z","shell.execute_reply":"2022-03-22T05:31:12.152923Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cust_sex1 = pd.DataFrame(cust_sex[[\"attribute\"]]).reset_index()\ncust_sex1[\"attribute\"] = cust_sex1[\"attribute\"].apply(\",\".join).astype(str)\ndel cust_sex\ncust_sex1","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:12.155483Z","iopub.execute_input":"2022-03-22T05:31:12.155722Z","iopub.status.idle":"2022-03-22T05:31:12.592059Z","shell.execute_reply.started":"2022-03-22T05:31:12.155693Z","shell.execute_reply":"2022-03-22T05:31:12.590895Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(cust_sex1.attribute.unique())","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:12.594943Z","iopub.execute_input":"2022-03-22T05:31:12.595292Z","iopub.status.idle":"2022-03-22T05:31:12.703723Z","shell.execute_reply.started":"2022-03-22T05:31:12.595246Z","shell.execute_reply":"2022-03-22T05:31:12.702803Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cust_sex1.loc[~((cust_sex1[\"attribute\"] == \"Woman\") |\n               (cust_sex1[\"attribute\"] == \"Young\")  |\n               (cust_sex1[\"attribute\"] == \"Man\")    |\n               (cust_sex1[\"attribute\"] == \"Have-kids\") |\n               (cust_sex1[\"attribute\"] == \"Sports-person\")), \"attribute\"] = \"Woman\"\ncust_sex1","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:12.705129Z","iopub.execute_input":"2022-03-22T05:31:12.705375Z","iopub.status.idle":"2022-03-22T05:31:13.749281Z","shell.execute_reply.started":"2022-03-22T05:31:12.705344Z","shell.execute_reply":"2022-03-22T05:31:13.748415Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(cust_sex1.attribute.unique())","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:13.750603Z","iopub.execute_input":"2022-03-22T05:31:13.750864Z","iopub.status.idle":"2022-03-22T05:31:13.85442Z","shell.execute_reply.started":"2022-03-22T05:31:13.750818Z","shell.execute_reply":"2022-03-22T05:31:13.853333Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"temp = cust_sex1[\"attribute\"].value_counts().index.to_list()\nplt.figure(figsize=(5, 5))\nplt.rcParams[\"font.size\"] = 12\nplt.pie(cust_sex1[\"attribute\"].value_counts().sort_values(ascending=False), \n        labels = temp, startangle = 90, counterclock=False, autopct=\"%1.1f%%\")\nplt.show","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:13.855968Z","iopub.execute_input":"2022-03-22T05:31:13.856243Z","iopub.status.idle":"2022-03-22T05:31:14.316994Z","shell.execute_reply.started":"2022-03-22T05:31:13.85621Z","shell.execute_reply":"2022-03-22T05:31:14.315641Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(cust_sex1[\"attribute\"].value_counts().sort_values(ascending=False))","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:14.319082Z","iopub.execute_input":"2022-03-22T05:31:14.319419Z","iopub.status.idle":"2022-03-22T05:31:14.520028Z","shell.execute_reply.started":"2022-03-22T05:31:14.319371Z","shell.execute_reply":"2022-03-22T05:31:14.518644Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df = pd.merge(customers_df, cust_sex1, on =\"customer_id\", how=\"left\")\ncustomers_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:14.522896Z","iopub.execute_input":"2022-03-22T05:31:14.523354Z","iopub.status.idle":"2022-03-22T05:31:17.178201Z","shell.execute_reply.started":"2022-03-22T05:31:14.523304Z","shell.execute_reply":"2022-03-22T05:31:17.17711Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df.isnull().sum()","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:17.181033Z","iopub.execute_input":"2022-03-22T05:31:17.181323Z","iopub.status.idle":"2022-03-22T05:31:17.845205Z","shell.execute_reply.started":"2022-03-22T05:31:17.181292Z","shell.execute_reply":"2022-03-22T05:31:17.844036Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df[\"attribute\"].fillna(\"Woman\", inplace = True)","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:17.846621Z","iopub.execute_input":"2022-03-22T05:31:17.84696Z","iopub.status.idle":"2022-03-22T05:31:18.003234Z","shell.execute_reply.started":"2022-03-22T05:31:17.846916Z","shell.execute_reply":"2022-03-22T05:31:18.002181Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"age_mean = customers_df[[\"age\", \"attribute\"]].groupby(\"attribute\").mean().round().reset_index()\nage_mean.columns = [\"attribute\", \"age_mean\"]\nage_mean","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:18.004726Z","iopub.execute_input":"2022-03-22T05:31:18.005091Z","iopub.status.idle":"2022-03-22T05:31:18.846579Z","shell.execute_reply.started":"2022-03-22T05:31:18.005054Z","shell.execute_reply":"2022-03-22T05:31:18.845767Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"customers_df = pd.merge(customers_df, age_mean, on = \"attribute\", how =\"left\")\ncustomers_df.loc[(customers_df[\"age\"].isnull()), \"age\"] = customers_df[\"age_mean\"]\ncustomers_df = customers_df.drop([\"age_mean\", \"age_id\"], axis =1)\ncustomers_df = pd.merge(customers_df, age_group, on=\"age\", how=\"left\")\ncustomers_df.isnull().sum()","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:18.847769Z","iopub.execute_input":"2022-03-22T05:31:18.848013Z","iopub.status.idle":"2022-03-22T05:31:21.565277Z","shell.execute_reply.started":"2022-03-22T05:31:18.847982Z","shell.execute_reply":"2022-03-22T05:31:21.564067Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del cust_sex1\ndel post_df\ndel custom_temp\ndel article_temp\ndel trans_temp","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:21.566632Z","iopub.execute_input":"2022-03-22T05:31:21.56688Z","iopub.status.idle":"2022-03-22T05:31:21.640888Z","shell.execute_reply.started":"2022-03-22T05:31:21.566836Z","shell.execute_reply":"2022-03-22T05:31:21.640039Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"transactions_df = pd.merge(transactions_df, customers_df, on =\"customer_id\", how =\"left\")\ntransactions_df.isnull().sum()","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:31:21.64228Z","iopub.execute_input":"2022-03-22T05:31:21.643325Z","iopub.status.idle":"2022-03-22T05:32:20.662854Z","shell.execute_reply.started":"2022-03-22T05:31:21.643243Z","shell.execute_reply":"2022-03-22T05:32:20.661926Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 6. Make prediction\n\nIn addition to \"age_id\" and \"sex/attribute\", region factor was also added. Therefore, some group have only a couple of items for recommendation. For those groups, the items derived from only age and sex/attribute factors (=recommend_2) will be recommended.\n\n前回の「年齢階層」「性別/属性」に加えて「地域」もグルーピング要素に加えたため、グループによっては売り上げ品目が２・３個に満たないグループが発生。こうしたグループには、「年齢階層」「性別/属性」のみのグルーピングによる売り上げ上位品目(recommend_2)を適用。","metadata":{}},{"cell_type":"code","source":"trans_df = transactions_df.loc[transactions_df.t_dat >= pd.to_datetime('2020-09-15')].copy() # changed from 2020-09-01\ntrans_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:20.664304Z","iopub.execute_input":"2022-03-22T05:32:20.664576Z","iopub.status.idle":"2022-03-22T05:32:33.596154Z","shell.execute_reply.started":"2022-03-22T05:32:20.664543Z","shell.execute_reply":"2022-03-22T05:32:33.595091Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"trans_df.article_id = ' ' + trans_df.article_id.astype('str')\ntemp = trans_df.groupby(['age_id','attribute','season','article_id'])['customer_id'].agg('count').reset_index()\ntemp.columns = ['age_id','attribute','season','article_id','count']\ntrans_df = trans_df.merge(temp, on=['age_id','attribute','season','article_id'], how='left')\ntrans_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:33.597761Z","iopub.execute_input":"2022-03-22T05:32:33.598534Z","iopub.status.idle":"2022-03-22T05:32:34.368458Z","shell.execute_reply.started":"2022-03-22T05:32:33.59849Z","shell.execute_reply":"2022-03-22T05:32:34.367411Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"trans_df = trans_df.sort_values(['count','t_dat'],ascending=False)\ntrans_df = trans_df.drop_duplicates(['age_id','attribute','season','article_id'])\ntrans_df","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:34.370203Z","iopub.execute_input":"2022-03-22T05:32:34.370592Z","iopub.status.idle":"2022-03-22T05:32:34.694848Z","shell.execute_reply.started":"2022-03-22T05:32:34.370542Z","shell.execute_reply":"2022-03-22T05:32:34.693999Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"recommend_2 = pd.DataFrame(trans_df.groupby(['age_id','attribute']).article_id.sum().reset_index())\nrecommend_2[\"len\"] = recommend_2[\"article_id\"].apply(lambda x : len(x))\nrecommend_2","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:34.696049Z","iopub.execute_input":"2022-03-22T05:32:34.69626Z","iopub.status.idle":"2022-03-22T05:32:34.966964Z","shell.execute_reply.started":"2022-03-22T05:32:34.696234Z","shell.execute_reply":"2022-03-22T05:32:34.965981Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"As you see below, for example, only 32 items ware sold for age_id=0(16-17 years old) Sports-person who live in warm area (sweater/coat is not sold). For those layers, the items based on age and sex/atrribute only will be recommended. \n\n例えば、age_idが0（16歳・17歳）でセーター・コートが売れていない地域に住むスポーツパーソンでは、9月に入って品目が2個しか売れていない。このような顧客階層には、地域要素を除いた売れ行き品目を適用する。","metadata":{}},{"cell_type":"code","source":"recommend_1 = pd.DataFrame(trans_df.groupby([\"age_id\",'attribute',\"season\"]).article_id.sum().reset_index())\nrecommend_1[\"len\"] = recommend_1[\"article_id\"].apply(lambda x : len(x))\nrecommend_1","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:34.968104Z","iopub.execute_input":"2022-03-22T05:32:34.968318Z","iopub.status.idle":"2022-03-22T05:32:35.14762Z","shell.execute_reply.started":"2022-03-22T05:32:34.968291Z","shell.execute_reply":"2022-03-22T05:32:35.146857Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"recommend_1 = pd.merge(recommend_1, recommend_2, on = [\"age_id\",\"attribute\"], how =\"left\")\nrecommend_1","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:35.148961Z","iopub.execute_input":"2022-03-22T05:32:35.149189Z","iopub.status.idle":"2022-03-22T05:32:35.175184Z","shell.execute_reply.started":"2022-03-22T05:32:35.149159Z","shell.execute_reply":"2022-03-22T05:32:35.173986Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"recommend_1[\"article_id\"] = recommend_1.loc[recommend_1[\"len_x\"]<131, \"article_id_y\"]\nrecommend_1[\"article_id\"].fillna(recommend_1[\"article_id_x\"], inplace =True)\nrecommend_1 = recommend_1.drop([\"article_id_x\",\"len_x\",\"article_id_y\",\"len_y\"], axis=1)\nrecommend_1","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:35.176602Z","iopub.execute_input":"2022-03-22T05:32:35.176884Z","iopub.status.idle":"2022-03-22T05:32:35.203193Z","shell.execute_reply.started":"2022-03-22T05:32:35.176824Z","shell.execute_reply":"2022-03-22T05:32:35.201608Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"recommend_1[\"article_id\"] = recommend_1[\"article_id\"].str.strip()\nrecommend_1[\"article_id\"] = recommend_1[\"article_id\"].str[:131]","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:35.204905Z","iopub.execute_input":"2022-03-22T05:32:35.205267Z","iopub.status.idle":"2022-03-22T05:32:35.213916Z","shell.execute_reply.started":"2022-03-22T05:32:35.205219Z","shell.execute_reply":"2022-03-22T05:32:35.2127Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission = pd.read_csv('../input/h-and-m-personalized-fashion-recommendations/sample_submission.csv')\nsubmission = submission[['customer_id']]\nsubmission = pd.merge(submission, customers_df, on = \"customer_id\", how = \"left\")\nsubmission = pd.merge(submission, recommend_1, on = [\"age_id\",'attribute',\"season\"], how=\"left\")\nsubmission = submission.drop([\"age\", \"age_id\", \"attribute\", \"town\", \"season\"], axis =1)\nsubmission.columns = (\"customer_id\", \"prediction\")\nsubmission.to_csv(\"submission.csv\",index=False)\nsubmission","metadata":{"execution":{"iopub.status.busy":"2022-03-22T05:32:35.215446Z","iopub.execute_input":"2022-03-22T05:32:35.216302Z","iopub.status.idle":"2022-03-22T05:32:56.588064Z","shell.execute_reply.started":"2022-03-22T05:32:35.216227Z","shell.execute_reply":"2022-03-22T05:32:56.587355Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}