{"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":"code","source":"import pandas as pd\n\npth = \"../input/digital-turbine-auction-bid-price-prediction/\"\ntrain = pd.read_csv(pth + \"train_data.csv\")\ntest = pd.read_csv(pth + \"test_data.csv\")\n\ntrain.sort_values(by=[\"eventTimestamp\"],ascending=True,inplace=True)\nwin_bids = train.groupby(['deviceId','unitDisplayType']).agg({'winBid':['last','mean']}).reset_index()\nwin_bids.columns = ['deviceId','unitDisplayType','last_winBid','mean_winBid']\ntest = test.merge(win_bids,on=['deviceId','unitDisplayType'],how='left')\n\ntest['winBid'] = 0.5*test['mean_winBid'] + 0.5*test['last_winBid']\ntest['winBid'] = test[['sentPrice','winBid']].max(axis=1)\n\ntest[['deviceId','winBid']].to_csv(\"avg_of_last_and_mean_winning_bid_per_unitDisplayType.csv\",index=False)","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-10-03T16:23:53.837218Z","iopub.execute_input":"2022-10-03T16:23:53.837633Z","iopub.status.idle":"2022-10-03T16:24:18.301892Z","shell.execute_reply.started":"2022-10-03T16:23:53.837598Z","shell.execute_reply":"2022-10-03T16:24:18.300863Z"},"trusted":true},"execution_count":null,"outputs":[]}]}