{"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":"# Introduction\n\n<span style=\"color:brown;font-family:Verdana;\">This challenge serves as final project for the \"How to win a data science competition\" Coursera course. Unfortunately this course have been removed from Coursera now.<br>In this competition we need to predict total sales for every product and store in the next month. Its challenging time-series dataset consisting of daily sales data, provided by one of the largest Russian software firms - **1C Company**.</span>\n           ![sales-forecast-3.png](attachment:76a5d0d8-f233-435d-8165-964ff325bd57.png) \n\n\n<span style=\"font-family:Verdana;\">**Files detail**: We have been provided below list of files,</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**sales_train**: the training set. Daily historical data from January 2013 to October 2015.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**test**: the test set. You need to forecast the sales for these shops and products for November 2015.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**items**: supplemental information about the items/products.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**item_categories**: supplemental information about the items categories.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**shops**: supplemental information about the shops.</span>\n\n<span style=\"font-family:Verdana;\">**Data fields**:</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**ID**: an Id that represents a (Shop, Item) tuple within the test set.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**shop_id**: unique identifier of a shop.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**item_id**: unique identifier of a product.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**item_category_id**: unique identifier of item category.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**item_cnt_day**: number of products sold. You are predicting a monthly amount of this measure.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**item_price**: current price of an item.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**date**: date in format dd/mm/yyyy.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**date_block_num** - a consecutive month number, used for convenience. January 2013 is 0, February 2013 is 1,..., October 2015 is 33.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**item_name**: name of item.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**shop_name**: name of shop.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">**item_category_name**: name of item category.</span>","metadata":{},"attachments":{"76a5d0d8-f233-435d-8165-964ff325bd57.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"# Data Extraction","metadata":{}},{"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom matplotlib.pyplot import figure\n%matplotlib inline \n\nimport tensorflow as tf\nfrom sklearn import preprocessing\n\nimport random\ntf.random.set_seed(53)\nrandom.seed(53)","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-05-04T09:27:59.678823Z","iopub.execute_input":"2022-05-04T09:27:59.680078Z","iopub.status.idle":"2022-05-04T09:27:59.689530Z","shell.execute_reply.started":"2022-05-04T09:27:59.680030Z","shell.execute_reply":"2022-05-04T09:27:59.688266Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"BASE=\"../input/competitive-data-science-predict-future-sales/\"\nitem_cat = pd.read_csv(BASE+\"item_categories.csv\")\nitem = pd.read_csv(BASE+\"items.csv\")\nsales_train = pd.read_csv(BASE+\"sales_train.csv\")\nshops = pd.read_csv(BASE+\"shops.csv\")\nsales_test = pd.read_csv(BASE+\"test.csv\")","metadata":{"execution":{"iopub.status.busy":"2022-05-04T09:28:02.456950Z","iopub.execute_input":"2022-05-04T09:28:02.457406Z","iopub.status.idle":"2022-05-04T09:28:03.775920Z","shell.execute_reply.started":"2022-05-04T09:28:02.457367Z","shell.execute_reply":"2022-05-04T09:28:03.775162Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def basic_eda(df):\n    print(\"----------TOP 5 RECORDS--------\")\n    print(df.head(5))\n    print(\"----------INFO-----------------\")\n    print(df.info())\n    print(\"----------Describe-------------\")\n    print(df.describe())\n    print(\"----------Columns--------------\")\n    print(df.columns)\n    print(\"----------Data Types-----------\")\n    print(df.dtypes)\n    print(\"-------Missing Values----------\")\n    print(df.isnull().sum())\n    print(\"-------NULL values-------------\")\n    print(df.isna().sum())\n    print(\"-----Shape Of Data-------------\")\n    print(df.shape)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T09:28:07.397321Z","iopub.execute_input":"2022-05-04T09:28:07.397733Z","iopub.status.idle":"2022-05-04T09:28:07.404622Z","shell.execute_reply.started":"2022-05-04T09:28:07.397704Z","shell.execute_reply":"2022-05-04T09:28:07.404011Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"=============================Sales Data=============================\")\nbasic_eda(sales_train)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T09:28:40.835568Z","iopub.execute_input":"2022-05-04T09:28:40.836602Z","iopub.status.idle":"2022-05-04T09:28:41.618328Z","shell.execute_reply.started":"2022-05-04T09:28:40.836555Z","shell.execute_reply":"2022-05-04T09:28:41.617341Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"=============================Test data=============================\")\nbasic_eda(sales_test)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T09:28:58.911754Z","iopub.execute_input":"2022-05-04T09:28:58.912149Z","iopub.status.idle":"2022-05-04T09:28:58.958714Z","shell.execute_reply.started":"2022-05-04T09:28:58.912107Z","shell.execute_reply":"2022-05-04T09:28:58.957596Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"=============================Item Categories=============================\")\nbasic_eda(item_cat)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T09:29:09.276624Z","iopub.execute_input":"2022-05-04T09:29:09.277015Z","iopub.status.idle":"2022-05-04T09:29:09.306637Z","shell.execute_reply.started":"2022-05-04T09:29:09.276975Z","shell.execute_reply":"2022-05-04T09:29:09.305444Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"=============================Items=============================\")\nbasic_eda(item)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T09:29:21.473261Z","iopub.execute_input":"2022-05-04T09:29:21.473579Z","iopub.status.idle":"2022-05-04T09:29:21.513960Z","shell.execute_reply.started":"2022-05-04T09:29:21.473546Z","shell.execute_reply":"2022-05-04T09:29:21.513033Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"=============================Shops=============================\")\nbasic_eda(shops)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T09:29:29.388219Z","iopub.execute_input":"2022-05-04T09:29:29.388524Z","iopub.status.idle":"2022-05-04T09:29:29.414222Z","shell.execute_reply.started":"2022-05-04T09:29:29.388479Z","shell.execute_reply":"2022-05-04T09:29:29.413568Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# EDA","metadata":{}},{"cell_type":"code","source":"corr = sales_train.corr()\ntop_corr_features = corr.index[abs(corr[\"item_cnt_day\"])>0]\n\nplt.figure(figsize=(6,6))\ng=sns.heatmap(sales_train[top_corr_features].corr(),annot=True,cmap=\"YlGnBu\")","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:43.593138Z","iopub.execute_input":"2022-05-04T08:47:43.593345Z","iopub.status.idle":"2022-05-04T08:47:44.657781Z","shell.execute_reply.started":"2022-05-04T08:47:43.593319Z","shell.execute_reply":"2022-05-04T08:47:44.656689Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sns.set(rc={'figure.figsize':(15,6)})\nsns.boxplot(x='shop_id', y='item_cnt_day', data=sales_train)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:44.659401Z","iopub.execute_input":"2022-05-04T08:47:44.659726Z","iopub.status.idle":"2022-05-04T08:47:48.493462Z","shell.execute_reply.started":"2022-05-04T08:47:44.659684Z","shell.execute_reply":"2022-05-04T08:47:48.492429Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sns.set(rc={'figure.figsize':(15,6)})\nsns.boxplot(x='date_block_num', y='item_cnt_day', data=sales_train)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:48.494685Z","iopub.execute_input":"2022-05-04T08:47:48.494932Z","iopub.status.idle":"2022-05-04T08:47:50.757860Z","shell.execute_reply.started":"2022-05-04T08:47:48.494904Z","shell.execute_reply":"2022-05-04T08:47:50.756914Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.scatter(sales_train['item_price'], sales_train['item_cnt_day'], color = \"red\")\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:50.761461Z","iopub.execute_input":"2022-05-04T08:47:50.761847Z","iopub.status.idle":"2022-05-04T08:47:56.008788Z","shell.execute_reply.started":"2022-05-04T08:47:50.761783Z","shell.execute_reply":"2022-05-04T08:47:56.007906Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df = sales_train.groupby(['date_block_num'],as_index=False).sum()\ndf.head()\nsns.set(rc={'figure.figsize':(15,6)})\nsns.lineplot(x='date_block_num', y='item_cnt_day', data=df)\nplt.axvline(x=11, ymin=0,ymax=1, color ='r')\nplt.axvline(x=23, ymin=0,ymax=1, color ='r')","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:56.010063Z","iopub.execute_input":"2022-05-04T08:47:56.010301Z","iopub.status.idle":"2022-05-04T08:47:56.533397Z","shell.execute_reply.started":"2022-05-04T08:47:56.010273Z","shell.execute_reply":"2022-05-04T08:47:56.532467Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_s = sales_train.groupby(['shop_id'],as_index=False).sum().sort_values(\"item_cnt_day\", ascending=False)\nsns.set(rc={'figure.figsize':(15,6)})\nsns.barplot(x='shop_id', y='item_cnt_day', data=df_s)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:56.534757Z","iopub.execute_input":"2022-05-04T08:47:56.535006Z","iopub.status.idle":"2022-05-04T08:47:58.580086Z","shell.execute_reply.started":"2022-05-04T08:47:56.534976Z","shell.execute_reply":"2022-05-04T08:47:58.579141Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_p = sales_train.groupby(['item_price'],as_index=False).sum()\nsns.set(rc={'figure.figsize':(15,6)})\nsns.lineplot(x='item_price', y='item_cnt_day', data=df_p)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:58.581458Z","iopub.execute_input":"2022-05-04T08:47:58.581718Z","iopub.status.idle":"2022-05-04T08:47:59.611340Z","shell.execute_reply.started":"2022-05-04T08:47:58.581685Z","shell.execute_reply":"2022-05-04T08:47:59.610301Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Feature Engineering\n\nLets start removing outliers first.","metadata":{}},{"cell_type":"code","source":"sales_train = sales_train[sales_train['item_price'] < 50000]","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:59.612754Z","iopub.execute_input":"2022-05-04T08:47:59.613243Z","iopub.status.idle":"2022-05-04T08:47:59.716926Z","shell.execute_reply.started":"2022-05-04T08:47:59.613208Z","shell.execute_reply":"2022-05-04T08:47:59.716061Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales_train = sales_train[sales_train['item_cnt_day'] < 1000]","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:59.718232Z","iopub.execute_input":"2022-05-04T08:47:59.718566Z","iopub.status.idle":"2022-05-04T08:47:59.831804Z","shell.execute_reply.started":"2022-05-04T08:47:59.718525Z","shell.execute_reply":"2022-05-04T08:47:59.830850Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(sum(item.duplicated(['item_name'])))\nprint(sum(item_cat.duplicated(['item_category_name'])))\nprint(sum(shops.duplicated(['shop_name'])))","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:59.833068Z","iopub.execute_input":"2022-05-04T08:47:59.833384Z","iopub.status.idle":"2022-05-04T08:47:59.857816Z","shell.execute_reply.started":"2022-05-04T08:47:59.833341Z","shell.execute_reply":"2022-05-04T08:47:59.856714Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"new_shop_id = {11: 10, 0: 57, 1: 58}\nshops['shop_id'] = shops['shop_id'].apply(lambda x: new_shop_id[x] if x in new_shop_id.keys() else x)\nsales_train['shop_id'] = sales_train['shop_id'].apply(lambda x: new_shop_id[x] if x in new_shop_id.keys() else x)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:47:59.859089Z","iopub.execute_input":"2022-05-04T08:47:59.859322Z","iopub.status.idle":"2022-05-04T08:48:00.994408Z","shell.execute_reply.started":"2022-05-04T08:47:59.859294Z","shell.execute_reply":"2022-05-04T08:48:00.993264Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales = pd.merge(sales_test, sales_train, on = ('shop_id', 'item_id'), how = 'left')","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:00.996769Z","iopub.execute_input":"2022-05-04T08:48:00.997160Z","iopub.status.idle":"2022-05-04T08:48:01.504297Z","shell.execute_reply.started":"2022-05-04T08:48:00.997113Z","shell.execute_reply":"2022-05-04T08:48:01.503039Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def missing_percent_of_column(train_set):\n    nan_percent = 100*(train_set.isnull().sum()/len(train_set))\n    nan_percent = nan_percent[nan_percent>0].sort_values(ascending=False).round(1)\n    DataFrame = pd.DataFrame(nan_percent)\n    # Rename the columns\n    mis_percent_table = DataFrame.rename(columns = {0 : '% of Misiing Values'}) \n    # Sort the table by percentage of missing descending\n    mis_percent = mis_percent_table\n    return mis_percent","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print('Missing sales details:')\nmiss_sales = missing_percent_of_column(sales)\nmiss_sales","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.505795Z","iopub.execute_input":"2022-05-04T08:48:01.506187Z","iopub.status.idle":"2022-05-04T08:48:01.514822Z","shell.execute_reply.started":"2022-05-04T08:48:01.506142Z","shell.execute_reply":"2022-05-04T08:48:01.512137Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# lets fill all NaN values with 0\nsales.fillna(0,inplace = True)\n# lets check our data now \nsales.head()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:49:09.444947Z","iopub.execute_input":"2022-05-04T08:49:09.445451Z","iopub.status.idle":"2022-05-04T08:49:09.552431Z","shell.execute_reply.started":"2022-05-04T08:49:09.445403Z","shell.execute_reply":"2022-05-04T08:49:09.551550Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.518952Z","iopub.status.idle":"2022-05-04T08:48:01.519319Z","shell.execute_reply.started":"2022-05-04T08:48:01.519139Z","shell.execute_reply":"2022-05-04T08:48:01.519159Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(sum(sales.duplicated()))","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.521394Z","iopub.status.idle":"2022-05-04T08:48:01.522243Z","shell.execute_reply.started":"2022-05-04T08:48:01.521967Z","shell.execute_reply":"2022-05-04T08:48:01.522002Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"💡 <span style=\"font-family:Verdana;\">We have duplicate records added after merging data, so we are handling duplicate record below way,</span>\n1. <span style=\"color:brown;font-family:Verdana;\">Remove all duplicate records for all column level.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">We will take column **date, date_block_num, shop_id, item_id, item_cnt_day** as unique records, and delete duplicate records.</span>\n1. <span style=\"color:brown;font-family:Verdana;\">Now lets take column **ID, date, date_block_num** as unique also, and delete duplicate records.</span>","metadata":{}},{"cell_type":"code","source":"sales = sales.drop_duplicates()\nsales = sales.drop_duplicates(['date','date_block_num','shop_id','item_id','item_cnt_day'])\nsales = sales.drop_duplicates(['ID','date','date_block_num'])\nsales.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.523240Z","iopub.status.idle":"2022-05-04T08:48:01.523600Z","shell.execute_reply.started":"2022-05-04T08:48:01.523415Z","shell.execute_reply":"2022-05-04T08:48:01.523444Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales.describe()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.528587Z","iopub.status.idle":"2022-05-04T08:48:01.528934Z","shell.execute_reply.started":"2022-05-04T08:48:01.528762Z","shell.execute_reply":"2022-05-04T08:48:01.528778Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales.loc[sales.item_cnt_day < 0, 'item_cnt_day'] = -1. * sales.loc[sales.item_cnt_day < 0, 'item_cnt_day']","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.530148Z","iopub.status.idle":"2022-05-04T08:48:01.530457Z","shell.execute_reply.started":"2022-05-04T08:48:01.530292Z","shell.execute_reply":"2022-05-04T08:48:01.530308Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales_month = sales.sort_values('date_block_num').groupby(['ID', 'date_block_num'], as_index = False).agg({'item_cnt_day': ['sum']})\nsales_month.columns = ['ID', 'date_block_num', 'item_cnt_month']\nsales_month.sample(10)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.531600Z","iopub.status.idle":"2022-05-04T08:48:01.532162Z","shell.execute_reply.started":"2022-05-04T08:48:01.531962Z","shell.execute_reply":"2022-05-04T08:48:01.531991Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales_month.describe()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.533208Z","iopub.status.idle":"2022-05-04T08:48:01.533929Z","shell.execute_reply.started":"2022-05-04T08:48:01.533570Z","shell.execute_reply":"2022-05-04T08:48:01.533589Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def to_IDs(np_data, col_ID):\n    # np_data - sales converted to numpy array\n    # col_ID - name of ID column\n    sales_by_ID = list()\n    IDs = np.unique(np_data[:,col_ID]).astype(int)\n    for i in IDs:\n        positions = np_data[:,col_ID] == i\n        sales_ID = np_data[positions,1:]\n        sales_by_ID.append(sales_ID)\n    return sales_by_ID, IDs","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.535109Z","iopub.status.idle":"2022-05-04T08:48:01.535651Z","shell.execute_reply.started":"2022-05-04T08:48:01.535471Z","shell.execute_reply":"2022-05-04T08:48:01.535491Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sales_by_id, id_list = to_IDs(sales_month.values,0)\nprint(len(sales_by_id))","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.536670Z","iopub.status.idle":"2022-05-04T08:48:01.537253Z","shell.execute_reply.started":"2022-05-04T08:48:01.537074Z","shell.execute_reply":"2022-05-04T08:48:01.537094Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# to decrease calculation time during a code debugging we remove IDs that don't have observtions for last months\ndef remove_ID_nan_last_year(np_data):\n    N_IDs = len(np_data)\n    col_date = 0\n    clear_data = list()\n    cut_month = 33 - 2\n    for i in range(N_IDs):\n        ID_data = np_data[i]\n        if len(ID_data[ID_data[:,col_date] >= cut_month,1]) != 0:\n            clear_data.append(ID_data)\n    return clear_data","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.538402Z","iopub.status.idle":"2022-05-04T08:48:01.538758Z","shell.execute_reply.started":"2022-05-04T08:48:01.538579Z","shell.execute_reply":"2022-05-04T08:48:01.538604Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's fill the missing date_block_num by NaN for paticular ID\ndef missing_months(np_data, col_date, col_TS, N_months = 34):\n    # col_date - index of date_block_num column\n    # col_TS - index of item_price column and item_cnt_month column\n    # at first fill time series by NaN for all months\n    series = [np.nan for _ in range(N_months)]\n    for i in range(len(np_data)):\n        position = int(np_data[i, col_date] - 1)\n        # fill positions that present in data\n        series[position] = np_data[i, col_TS]\n    return series","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.540024Z","iopub.status.idle":"2022-05-04T08:48:01.540776Z","shell.execute_reply.started":"2022-05-04T08:48:01.540553Z","shell.execute_reply":"2022-05-04T08:48:01.540581Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's fill the missing item_cnt_month and item_price for particular ID\ndef to_fill_missing(np_data, N_months = 34):\n    col = ['date_block_num','item_cnt_month']\n    sales_ID = pd.DataFrame(np_data, columns = col)\n    if sales_ID.shape[0] < N_months:\n        date_month = pd.DataFrame(range(N_months),columns = ['date_block_num'])\n        sales_ID = pd.merge(date_month, sales_ID, on = ('date_block_num'), how = 'left')\n        sales_ID = sales_ID.reindex(columns = col)\n        sales_ID['item_cnt_month'] = sales_ID['item_cnt_month'].fillna(0.0)\n    return sales_ID['item_cnt_month'].to_numpy()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.542141Z","iopub.status.idle":"2022-05-04T08:48:01.542469Z","shell.execute_reply.started":"2022-05-04T08:48:01.542297Z","shell.execute_reply":"2022-05-04T08:48:01.542321Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Plot time series for particular ID to find out missing months\ndef plot_TS(np_data, n_vars = 1, N_months = 34, flag = 0):\n    # n_vars = 1 or 2 (plot item_cnt OR item_cnt and item_price)\n    plt.figure()\n    if flag == 1:\n        TSs = to_fill_missing(np_data, N_months)\n    for i in range(n_vars):\n        col_plot = i + 1 # index of column to plot\n        if flag == 1:\n            series = TSs#[:,col_plot]\n        else:\n            series = missing_months(np_data, 0, col_plot, N_months)\n        ax = plt.subplot(n_vars, 1, i+1)\n        plt.plot(series, 'o')\n        plt.plot(series)\n    plt.show()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.543769Z","iopub.status.idle":"2022-05-04T08:48:01.544351Z","shell.execute_reply.started":"2022-05-04T08:48:01.544145Z","shell.execute_reply":"2022-05-04T08:48:01.544169Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for i in np.random.randint(0, len(sales_by_id), 5):\n    plot_TS(sales_by_id[i], flag = 1)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.545885Z","iopub.status.idle":"2022-05-04T08:48:01.546403Z","shell.execute_reply.started":"2022-05-04T08:48:01.546192Z","shell.execute_reply":"2022-05-04T08:48:01.546219Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's create 2D-array and each column is counts of particular ID where missing months is filled\ndef full_data(data, N_months = 34):\n    N_IDs = len(data)\n    TS = np.empty((N_months, N_IDs))\n    for i in range(N_IDs):\n        TS[:, i] = to_fill_missing(data[i], N_months)\n    return TS","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.547289Z","iopub.status.idle":"2022-05-04T08:48:01.547941Z","shell.execute_reply.started":"2022-05-04T08:48:01.547699Z","shell.execute_reply":"2022-05-04T08:48:01.547724Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"TS = full_data(sales_by_id)\nTS.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.549219Z","iopub.status.idle":"2022-05-04T08:48:01.549539Z","shell.execute_reply.started":"2022-05-04T08:48:01.549373Z","shell.execute_reply":"2022-05-04T08:48:01.549394Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"val_month = 29\nvalid_TS = TS[val_month:,:]\ntrain_TS = TS[:val_month,:]\n\nprint(train_TS.shape, valid_TS.shape)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.551210Z","iopub.status.idle":"2022-05-04T08:48:01.551899Z","shell.execute_reply.started":"2022-05-04T08:48:01.551692Z","shell.execute_reply":"2022-05-04T08:48:01.551714Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"scaler = preprocessing.MinMaxScaler()\nscaler.fit(TS)\n\ntrain_scaled = scaler.transform(train_TS)\nvalid_scaled = scaler.transform(valid_TS)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.552975Z","iopub.status.idle":"2022-05-04T08:48:01.553478Z","shell.execute_reply.started":"2022-05-04T08:48:01.553306Z","shell.execute_reply":"2022-05-04T08:48:01.553326Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Training","metadata":{}},{"cell_type":"code","source":"def to_make_features(TS, n_lag, batch_size):\n    ds = tf.data.Dataset.from_tensor_slices(TS) # each element of dataset is one value of TS \n    ds = ds.window(n_lag+1, shift = 1, drop_remainder = True) # (n_lag+1)-elements of dataset is combined to window\n    ds = ds.flat_map(lambda row: row.batch(n_lag + 1)) # to batch elements in window to tensor (one element) and to flat (now there are no windows)\n    # Let's shuffle befor we combine batches for epoch\n    ds = ds.shuffle(300)\n    # make the tuple: first element is features, second element is labels\n    # features-(1,2,3) and labels-(2,3,4). 2 goes after 1, 3 goes after 2, 4 goes after 3.\n    ds = ds.map(lambda row: (row[:-1,:], row[1:,:]))\n    # combine tuples to banch for gradient descent\n    # instead of a row we will have a matrix in every tuple\n    ds = ds.batch(batch_size).prefetch(1)\n    return ds","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.554604Z","iopub.status.idle":"2022-05-04T08:48:01.555171Z","shell.execute_reply.started":"2022-05-04T08:48:01.554935Z","shell.execute_reply":"2022-05-04T08:48:01.554965Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"n_lag = 6\nbatch_size = 8\nfeatures = to_make_features(train_scaled, n_lag, batch_size)\nval_features = to_make_features(valid_scaled, n_lag, batch_size)\nConv_filters = 64\nConv_kernel_size = 4\nLSTM_filters = 64\nn_outputs = train_scaled.shape[1]","metadata":{"_kg_hide-output":true,"execution":{"iopub.status.busy":"2022-05-04T08:48:01.556103Z","iopub.status.idle":"2022-05-04T08:48:01.556428Z","shell.execute_reply.started":"2022-05-04T08:48:01.556259Z","shell.execute_reply":"2022-05-04T08:48:01.556280Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model = tf.keras.models.Sequential([\n  tf.keras.layers.Conv1D(filters = Conv_filters, kernel_size = Conv_kernel_size,\n                      strides=1, padding=\"causal\", activation=\"relu\", input_shape=[None, n_outputs]),\n  tf.keras.layers.Dropout(rate=0.2),\n  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(LSTM_filters, return_sequences=True)),\n  tf.keras.layers.Dropout(rate=0.2),\n  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(LSTM_filters, return_sequences=True)),\n  tf.keras.layers.Dropout(rate=0.2),\n    \n  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(LSTM_filters, return_sequences=True)),\n  tf.keras.layers.Dense(n_outputs)])\n\nlr_schedule = tf.keras.callbacks.LearningRateScheduler(lambda epoch: 1e-5 * 10**(epoch / 20))\noptimizer = tf.keras.optimizers.Adam(learning_rate = 1e-5)\n\nmodel.compile(loss=tf.keras.losses.Huber(), optimizer=optimizer, metrics=[\"mean_squared_error\"])\nmodel.summary()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.557804Z","iopub.status.idle":"2022-05-04T08:48:01.558348Z","shell.execute_reply.started":"2022-05-04T08:48:01.558146Z","shell.execute_reply":"2022-05-04T08:48:01.558168Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fitting = model.fit(features, epochs=80,verbose=0, callbacks=[lr_schedule])\nplt.semilogx(fitting.history[\"lr\"], fitting.history[\"loss\"])","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.559624Z","iopub.status.idle":"2022-05-04T08:48:01.560271Z","shell.execute_reply.started":"2022-05-04T08:48:01.560042Z","shell.execute_reply":"2022-05-04T08:48:01.560070Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"optimizer = tf.keras.optimizers.Adam(learning_rate = 1e-2)\nmodel.compile(loss=tf.keras.losses.Huber(), optimizer=optimizer, metrics=[\"mae\"])\nfitting = model.fit(features, epochs=300, verbose = 0, validation_data = val_features)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.561523Z","iopub.status.idle":"2022-05-04T08:48:01.561893Z","shell.execute_reply.started":"2022-05-04T08:48:01.561694Z","shell.execute_reply":"2022-05-04T08:48:01.561716Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def model_forecast(model, TS, n_lag, batch_size):\n    ds = tf.data.Dataset.from_tensor_slices(TS)\n    ds = ds.window(n_lag, shift=1, drop_remainder=True)\n    ds = ds.flat_map(lambda row: row.batch(n_lag))\n    ds = ds.batch(batch_size)\n    forecast = model.predict(ds)\n    return forecast","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.563281Z","iopub.status.idle":"2022-05-04T08:48:01.563635Z","shell.execute_reply.started":"2022-05-04T08:48:01.563457Z","shell.execute_reply":"2022-05-04T08:48:01.563481Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"forecast = model_forecast(model, train_scaled, n_lag, batch_size)\nforecast = forecast[:,-1,:]\nforecast = scaler.inverse_transform(forecast)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.565538Z","iopub.status.idle":"2022-05-04T08:48:01.566082Z","shell.execute_reply.started":"2022-05-04T08:48:01.565842Z","shell.execute_reply":"2022-05-04T08:48:01.565872Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"iplot = 0\nfor i in np.random.randint(0, n_outputs, 4):\n    iplot += 1\n    plt.subplot(4,1,iplot)\n    plt.plot(range(n_lag, val_month+1), np.append(train_TS[n_lag:,i],valid_TS[0,i]), 'r')\n    plt.plot(range(n_lag, val_month+1), forecast[:,i], 'b')\n    plt.legend([\"actual\", \"predicted\"])","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.566933Z","iopub.status.idle":"2022-05-04T08:48:01.567598Z","shell.execute_reply.started":"2022-05-04T08:48:01.567380Z","shell.execute_reply":"2022-05-04T08:48:01.567407Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ds = tf.data.Dataset.from_tensor_slices(TS)\nds = ds.window(n_lag, shift=1, drop_remainder=True)\nds = ds.flat_map(lambda row: row.batch(n_lag))\nds = ds.batch(batch_size)\npredict = model.predict(ds)\nlast_month_predict = predict[-1,-1,:]\nlast_month_forecast = scaler.inverse_transform(np.expand_dims(last_month_predict, axis = 0))","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.568859Z","iopub.status.idle":"2022-05-04T08:48:01.569380Z","shell.execute_reply.started":"2022-05-04T08:48:01.569191Z","shell.execute_reply":"2022-05-04T08:48:01.569217Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission = pd.DataFrame({'ID': id_list, 'item_cnt_month': np.squeeze(last_month_predict) })\nsubmission.loc[submission.item_cnt_month < 0, 'item_cnt_month'] = 0\nsubmission = pd.merge(sales_test.ID, submission, on = ('ID'), how = 'left')\nsubmission = submission.fillna(0)\nsubmission.head()","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.574309Z","iopub.status.idle":"2022-05-04T08:48:01.574819Z","shell.execute_reply.started":"2022-05-04T08:48:01.574640Z","shell.execute_reply":"2022-05-04T08:48:01.574661Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission.to_csv('submission.csv', index=False)","metadata":{"execution":{"iopub.status.busy":"2022-05-04T08:48:01.576050Z","iopub.status.idle":"2022-05-04T08:48:01.576585Z","shell.execute_reply.started":"2022-05-04T08:48:01.576407Z","shell.execute_reply":"2022-05-04T08:48:01.576428Z"},"trusted":true},"execution_count":null,"outputs":[]}]}