{"cells":[{"metadata":{},"cell_type":"markdown","source":"![image](https://www.technotification.com/wp-content/uploads/2018/09/Renewable-Energy-Ideas-1200x600.jpg)\nImage source: [technotification](https://www.technotification.com/2018/09/amazing-renewable-energy-ideas.html)"},{"metadata":{},"cell_type":"markdown","source":"This notebook aims to predict a building's energy consumption over 2017 and 2018 using the data from 2016 in 4 different consumpiton categories (electricity, chilled water, steam, hot water) using ASHRAE data, which is our problem statement as well.\n\nThis is a supervised machine learning model, meaning based on the columns available in the datasets and data from 2016, we are going to train the model to predict an energy consumption of a building in each category. Since, consumption values are labeled as meter_reading and they are continuous, we are going to apply regression techniques to generate predictions on meter_reading.\n\nIt is a highly debated and popular competition in Kaggle currently, however my main motivation is to contribute to make energy-efficient buildings by estimating its energy consumption. It seemed like a good start to save our energy for future!\n\nThere will be 3 notebooks covering the complete machine learning building pipeline. \n* [Notebook 1](https://www.kaggle.com/cereniyim/save-the-energy-for-the-future-1-detailed-eda) covered understanding, formating the data and detailed EDA. \n* [Notebook 2](https://www.kaggle.com/cereniyim/save-the-energy-for-the-future-2-fe-lightgbm) consists of the feature engineering, building the model and tuning its parameters.\n\nThis third and last notebook will focus on parts 6, 7 and 8 focusing on generating predictions, evaluating them and a summary about the whole project.\n\n1) Understand, Cleand and Format Data\n\n2) Exploratory Data Analysis\n\n3) Feature Engineering & Selection\n\n4) Compare Several Machine Learning Models\n\n5) Perform Hyperparameter Tuning and Cross Validation\n\n**6) Evaluate Model with Test Data**\n\n**7) Interpret Model Results**\n\n**8) Summary & Conclusions**\n\nMachine Learning application and building is not a linear and one time process. Steps above enable me to follow a structured way for an end-to-end machine project flow and preparation for the each step ahead. All in all, steps might be modified or revisited according to findings. You can use the table of contents to navigate to each section. 👇\n\nEnjoy reading !"},{"metadata":{},"cell_type":"markdown","source":"# <a id = 'top'> Table of Contents </a>\n  - <a href = '#1'> Quick Recap from Previous Notebook </a>\n  - <a href = '#6'> 6. Evaluate Model with Test Data </a>\n    - <a href = '#6-1'> 6.1. Load training data for re-creating model </a>\n    - <a href = '#6-2'> 6.2. Build model </a>\n    - <a href = '#6-3'> 6.3. Feature alignment between training and test set & Imputation </a>\n      - <a href = '#6-3-1'> 6.3.1. Building features transformation </a>\n      - <a href = '#6-3-2'> 6.3.2. Weather features transformation </a>\n      - <a href = '#6-3-3'> 6.3.3. Create one test dataframe </a>\n      - <a href = '#6-3-4'> 6.3.4. Split test data for 2017 and 2018 </a>\n      - <a href = '#6-3-5'> 6.3.5. Impute missing values </a>\n    - <a href = '#6-4'> 6.4. Make sure test data shape and check for missing values </a>\n    - <a href = '#6-5'> 6.5. Generate predictions </a>\n  - <a href = '#7'> 7. Interpret Model Results </a>\n  - <a href = '#8'> 8. Submissions & Summary & Conclusions </a>"},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# for data manipulation\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# pandas options\npd.set_option('display.max_columns', 50)\npd.set_option('display.float_format', lambda x: '%.3f' % x)\npd.set_option('mode.use_inf_as_na', True)\npd.options.mode.chained_assignment = None\n\n# for date manipulation\nfrom datetime import datetime\n\n# for visualization: matplotlib\nfrom matplotlib import pyplot as plt\nfrom IPython.core.pylabtools import figsize\n%matplotlib inline\n# to display visuals in the notebook\n\n# for visualization: seaborn\nimport seaborn as sns\nsns.set_context(font_scale=2)\n\n# for data preprocessing\nfrom sklearn.preprocessing import LabelEncoder\nfrom itertools import combinations\nfrom sklearn.model_selection import KFold\n\n# for building the model and calculate RMSE\nimport lightgbm as lgb\nfrom sklearn.metrics import mean_squared_error\nfrom math import sqrt\n\n# to cleanup memory usage\nimport gc\n\n# Input data files are available in the \"../input/\" directory.\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# Any results you write to the current directory are saved as output.","execution_count":null,"outputs":[]},{"metadata":{"_kg_hide-input":true,"trusted":true,"_kg_hide-output":true},"cell_type":"code","source":"## Function to reduce the DF size and reduce test dataframe size\ndef reduce_memory_usage(df, verbose=True):\n    numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']\n    start_mem = df.memory_usage().sum() / 1024**2    \n    for col in df.columns:\n        col_type = df[col].dtypes\n        if col_type in numerics:\n            c_min = df[col].min()\n            c_max = df[col].max()\n            if str(col_type)[:3] == 'int':\n                if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:\n                    df[col] = df[col].astype(np.int8)\n                elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:\n                    df[col] = df[col].astype(np.int16)\n                elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:\n                    df[col] = df[col].astype(np.int32)\n                elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:\n                    df[col] = df[col].astype(np.int64)  \n            else:\n                if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:\n                    df[col] = df[col].astype(np.float16)\n                elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:\n                    df[col] = df[col].astype(np.float32)\n                else:\n                    df[col] = df[col].astype(np.float64)    \n    end_mem = df.memory_usage().sum() / 1024**2\n    if verbose: print('Mem. usage decreased to {:5.2f} Mb ({:.1f}% reduction)'.format(end_mem, 100 * (start_mem - end_mem) / start_mem))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# <a id = '1'> Quick Recap from Previous Notebook </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{},"cell_type":"markdown","source":"In the previous notebook, we focused on feature engineering, building model and improving the evaluation metric (RMSE) further with cross validation and hyperparameter optimization.\n\nWith feature engineering, we extracted and transformed features from the existing columns, added features by combining weather variables. Afterwards, with a univariate feature selection method, looking at the pearson coefficients of the features to the log1p of the meter_reading, we determined on final 24 features.\n\nFor model selection, linear regression, k nearest neighbor regressor and light gradient boosting models are built. Light gbm delivered the best results. After applying the cross-validation results RMSE improved from 1.21 to 1.1, hyperparameter search did not help much in improving the results. \n\n**Here is a visual from how training and validation error improved in from one of the cross-validation folds of Light GBM:**\n![image.png](attachment:image.png)","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"I will perform same steps performed in the feature engineering to align test set with the training set. With the cross validated model I will generate the predictions and evaluate them.\n\nFinally, I will end the project with an overall summary."},{"metadata":{},"cell_type":"markdown","source":"# <a id = '6'> 6. Evaluate Model with Test Data </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{},"cell_type":"markdown","source":"## <a id = '6-1'> 6.1. Load training data for re-creating model </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"# load training data created in the second notebook into dataframes\nX = pd.read_csv(\"/kaggle/input/save-the-energy-for-the-future-2-fe-lightgbm/X.csv\")\ny = pd.read_csv(\"/kaggle/input/save-the-energy-for-the-future-2-fe-lightgbm/y.csv\", header=None)\n\nreduce_memory_usage(X)\nreduce_memory_usage(y)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# rename target as log_meter_reading\ny.rename(columns = {0: \"log_meter_reading\"}, \n         inplace=True)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## <a id = '6-2'> 6.2. Build model </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create categorical features \ncategorical_features = ['building_id', 'site_id', 'meter',\n                        'primary_use', 'wind_compass_direction',\n                        'day_of_week', 'hour','is_weekend', 'season']\n\n# initial parameters of light gbm algorithm\ninitial_params = {\"objective\": \"regression\",\n                  \"boosting\": \"gbdt\",\n                  \"num_leaves\": 60,\n                  \"learning_rate\": 0.05,\n                  \"feature_fraction\": 0.85,\n                  \"reg_lambda\": 2,\n                  \"metric\": {'rmse'}\n}","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# cretae kfold object and empty model and evaluation lists\nkf = KFold(n_splits=4, shuffle=False, random_state=42)\n\n# save 4 model as a list\nmodels = []\n\n# dynamically split X and y with the k-fold split indexes\nfor train_index,valid_index in kf.split(X):\n    X_train_kf = X.loc[train_index]\n    y_train_kf = y.loc[train_index]\n    \n    X_valid_kf = X.loc[valid_index]\n    y_valid_kf = y.loc[valid_index]\n    \n    d_train = lgb.Dataset(X_train_kf, \n                          label=y_train_kf,\n                          categorical_feature=categorical_features, \n                          free_raw_data=False)\n    \n    d_valid = lgb.Dataset(X_valid_kf, \n                          label=y_valid_kf,\n                          categorical_feature=categorical_features, \n                          free_raw_data=False)\n    \n    model = lgb.train(initial_params, \n                      train_set=d_train, \n                      num_boost_round=1000, \n                      valid_sets=[d_train, d_valid],\n                      verbose_eval=250, \n                      early_stopping_rounds=500)\n    \n    models.append(model)\n    \n    del X_train_kf, y_train_kf, X_valid_kf, y_valid_kf, d_train, d_valid\n    gc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## <a id = '6-3'> 6.3. Feature alignment between training and test set & Imputation </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{},"cell_type":"markdown","source":"Let's have a look at the columns of X and align accordingly"},{"metadata":{"trusted":true},"cell_type":"code","source":"X.columns","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"del X\ndel y\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# load building  data modify\nbuilding = pd.read_csv(\"/kaggle/input/ashrae-energy-prediction/building_metadata.csv\")\n# drop floor_count\nbuilding.drop(columns=[\"floor_count\"], inplace=True)\n\n# load weather_test modify\nweather_test = pd.read_csv(\"/kaggle/input/ashrae-energy-prediction/weather_test.csv\")\nweather_test[\"timestamp\"] = pd.to_datetime(weather_test[\"timestamp\"],\n                                            format='%Y-%m-%d %H:%M:%S')\n\n# load test and modify\ntest = pd.read_csv(\"/kaggle/input/ashrae-energy-prediction/test.csv\")\ntest[\"timestamp\"] = pd.to_datetime(test[\"timestamp\"],\n                                   format='%Y-%m-%d %H:%M:%S')\n\nreduce_memory_usage(building)\nreduce_memory_usage(weather_test)\nreduce_memory_usage(test)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### <a id = '6-3-1'> 6.3.1. Building features transformation </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"# add building age column\ncurrent_year = datetime.now().year\nbuilding['building_age'] = current_year - building['year_built']\nbuilding.drop(columns=['year_built'], inplace=True)\n\n# since NA values only present in building age fillna can be used\nbuilding.fillna(round(building.building_age.mean(),0),\n                inplace=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# create label encoder object and transform the column\nle = LabelEncoder()\nle_primary_use = le.fit_transform(building.primary_use)\n\n# add label encoded column to dataframe\nbuilding['primary_use'] = le_primary_use\n\ndel le, le_primary_use\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# check if any NA values left\nbuilding.isna().sum()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### <a id = '6-3-2'> 6.3.2. Weather features transformation </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"def convert_season(month):\n    if (month <= 2) | (month == 12):\n        return 0\n    # as winter\n    elif month <= 5:\n        return 1\n    # as spring\n    elif month <= 8:\n        return 2\n    # as summer\n    elif month <= 11:\n        return 3\n    # as fall","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# add month, day of week, day of month, hour, season\nweather_test['month'] = weather_test['timestamp'].dt.month.astype(np.int8)\nweather_test['day_of_week'] = weather_test['timestamp'].dt.dayofweek.astype(np.int8)\nweather_test['day_of_month']= weather_test['timestamp'].dt.day.astype(np.int8)\nweather_test['hour'] = weather_test['timestamp'].dt.hour\nweather_test['season'] = weather_test.month.apply(convert_season)\n\n# add is weekend column\nweather_test['is_weekend'] = weather_test.day_of_week.apply(lambda x: 1 if x>=5 else 0)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Reset Index for Update**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# reset index for fast update\nweather_test = weather_test.set_index(\n    ['site_id','day_of_month','month'])","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Air Temperature**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create dataframe of daily means per site id\nair_temperature_filler = pd.DataFrame(weather_test\n                                      .groupby(['site_id','day_of_month','month'])\n                                      ['air_temperature'].mean(),\n                                      columns=[\"air_temperature\"])\n# create dataframe of air_temperatures to fill\ntemporary_df = pd.DataFrame({'air_temperature' : weather_test.air_temperature})\n\n# update NA air_temperature values\ntemporary_df.update(air_temperature_filler, overwrite=False)\n\n# update in the weather train dataset\nweather_test[\"air_temperature\"] = temporary_df[\"air_temperature\"]\n\ndel temporary_df, air_temperature_filler\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Cloud Coverage**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create dataframe of daily means per site id\ncloud_coverage_filler = pd.DataFrame(weather_test\n                                     .groupby(['site_id','day_of_month','month'])\n                                     ['cloud_coverage'].mean(),\n                                     columns = ['cloud_coverage'])\ncloud_coverage_filler.fillna(round(cloud_coverage_filler.cloud_coverage.mean(),0), \n                             inplace=True)\n\n# create dataframe of cloud_coverages to fill\ntemporary_df = pd.DataFrame({'cloud_coverage' : weather_test.cloud_coverage})\n\n# update NA cloud_coverage values\ntemporary_df.update(cloud_coverage_filler, overwrite=False)\n\n# update in the weather train dataset\nweather_test[\"cloud_coverage\"] = temporary_df[\"cloud_coverage\"]\n\ndel temporary_df, cloud_coverage_filler\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Dew Temperature**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create dataframe of daily means per site id\ndew_temperature_filler = pd.DataFrame(weather_test\n                                      .groupby(['site_id','day_of_month','month'])\n                                      ['dew_temperature'].mean(),\n                                      columns=[\"dew_temperature\"])\n# create dataframe of dew_temperatures to fill\ntemporary_df = pd.DataFrame({'dew_temperature' : weather_test.dew_temperature})\n\n# update NA dew_temperature values\ntemporary_df.update(dew_temperature_filler, overwrite=False)\n\n# update in the weather train dataset\nweather_test[\"dew_temperature\"] = temporary_df[\"dew_temperature\"]\n\ndel temporary_df, dew_temperature_filler\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Precip Depth 1 Hour**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create dataframe of daily means per site id\nprecip_depth_filler = pd.DataFrame(weather_test\n                                   .groupby(['site_id','day_of_month','month'])\n                                   ['precip_depth_1_hr'].mean(),\n                                   columns=['precip_depth_1_hr'])\nprecip_depth_filler.fillna(round(precip_depth_filler['precip_depth_1_hr'].mean(),0)\n                           , inplace=True)\n\n# create dataframe of precip_depth_1_hr to fill\ntemporary_df = pd.DataFrame({'precip_depth_1_hr' : weather_test.precip_depth_1_hr})\n\n# update NA precip_depth_1_hr values\ntemporary_df.update(precip_depth_filler, overwrite=False)\n\n# update in the weather train dataset\nweather_test[\"precip_depth_1_hr\"] = temporary_df[\"precip_depth_1_hr\"]\n\ndel precip_depth_filler, temporary_df\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Sea Level Pressure**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create dataframe of daily means per site id\nsea_level_filler = pd.DataFrame(weather_test\n                                .groupby(['site_id','day_of_month','month'])\n                                ['sea_level_pressure'].mean(),\n                                columns=['sea_level_pressure'])\nmean_sea_level_pressure = round(\n    sea_level_filler\n    ['sea_level_pressure']\n    .astype(float)\n    .mean(),2)\n\nsea_level_filler.fillna(mean_sea_level_pressure, inplace=True)\n\n# create dataframe of sea_level_pressure to fill\ntemporary_df = pd.DataFrame({'sea_level_pressure' : weather_test.sea_level_pressure})\n\n# update NA sea_level_pressure values\ntemporary_df.update(sea_level_filler, overwrite=False)\n\n# update in the weather train dataset\nweather_test[\"sea_level_pressure\"] = temporary_df[\"sea_level_pressure\"]\n\ndel sea_level_filler, temporary_df\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Wind Direction**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create dataframe of daily means per site id\nwind_direction_filler = pd.DataFrame(weather_test\n                                     .groupby(['site_id','day_of_month','month'])\n                                     ['wind_direction'].mean(),\n                                     columns=['wind_direction'])\n# create dataframe of wind_direction to fill\ntemporary_df = pd.DataFrame({'wind_direction' : weather_test.wind_direction})\n\n# update NA wind_direction values\ntemporary_df.update(wind_direction_filler, overwrite=False)\n\n# update in the weather train dataset\nweather_test[\"wind_direction\"] = temporary_df[\"wind_direction\"]\n\ndel temporary_df, wind_direction_filler\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Wind Speed**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create dataframe of daily means per site id\nwind_speed_filler = pd.DataFrame(weather_test\n                                 .groupby(['site_id','day_of_month','month'])\n                                 ['wind_speed'].mean(),\n                                 columns=['wind_speed'])\n# create dataframe of wind_speed to fill\ntemporary_df = pd.DataFrame({'wind_speed' : weather_test.wind_speed})\n\n# update NA wind_speed values\ntemporary_df.update(wind_speed_filler, overwrite=False)\n\n# update in the weather train dataset\nweather_test[\"wind_speed\"] = temporary_df[\"wind_speed\"]\n\ndel temporary_df, wind_speed_filler\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# check if NA values are left\nweather_test.isna().sum()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Reset indexes to transfrom weather dataframe to original form**"},{"metadata":{"trusted":true},"cell_type":"code","source":"weather_test = weather_test.reset_index()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Convert Wind Direction**"},{"metadata":{"trusted":true},"cell_type":"code","source":"def convert_direction(degrees):\n    if degrees <= 90:\n        return 0\n    # as norteast direction\n    elif degrees <= 180:\n        return 1\n    # as southeast direction\n    elif degrees <= 270:\n        return 2\n    # as southwest direction\n    elif degrees <= 360:\n        return 3\n    # as northwest direction\nweather_test['wind_compass_direction'] = weather_test.wind_direction.apply(convert_direction)\nweather_test.drop(columns=['wind_direction'], inplace=True)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**Add combination of weather variables**"},{"metadata":{"trusted":true},"cell_type":"code","source":"# create weather variables combinations\nweather_test['meansea_level_pressure_wind_speed'] = (weather_test['sea_level_pressure'] +\n                                                     weather_test['wind_speed']) / 2\nweather_test['meancloud_coverage_sea_level_pressure'] = (weather_test['sea_level_pressure'] + \n                                                         weather_test['cloud_coverage']) / 2\nweather_test['meancloud_coverage_wind_speed '] = (weather_test['cloud_coverage'] + \n                                                  weather_test['wind_speed']) / 2\nweather_test['meanprecip_depth_1_hr_sea_level_pressure'] = (weather_test['precip_depth_1_hr'] + \n                                                            weather_test['sea_level_pressure']) / 2\nweather_test['meanair_temperature_sea_level_pressure'] = (weather_test['air_temperature'] + \n                                                          weather_test['sea_level_pressure']) / 2","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### <a id = '6-3-3'> 6.3.3. Create one test dataframe </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"# merge dataframes on test dataframe\ntest = test.merge(building, on = \"building_id\", how = \"left\")\ntest = test.merge(weather_test, on = [\"site_id\", \"timestamp\"], how=\"left\")\n\n# delete the other ones to save space from the memory\ndel weather_test\ndel building\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test.columns","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test.drop(columns = [\"row_id\", \n                     \"timestamp\"], inplace=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#feature_set = ['building_age', 'le_primary_use', 'cloud_coverage',\n#               'is_weekend','wind_speed', 'day_of_week',\n#               'wind_compass_direction', 'sea_level_pressure', 'air_temperature',\n#               'day_of_month', 'dew_temperature', 'hour', \n#               'month', 'meter', 'building_id', \n#               'site_id', 'floor_count', 'square_feet', 'year']","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(\"Number of unique columns in the test dataset:\", test.shape[1])","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"With the memory usage concerns, I will split test dataset for 2017 and 2018 based on `year` column. If we ignore year column test dataset has 24 columns, aligning with the number of columns X has. Now let's check if there are any NA columns still exist after the merge:"},{"metadata":{"trusted":true},"cell_type":"code","source":"test.isna().sum()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### <a id = '6-3-4'> 6.3.4. Split test data for 2017 and 2018 </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"# split test set into two for faster imputations\nX_test_2017 = test[:20848800]\nX_test_2018 = test[20848800:]\n\ndel test\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### <a id = '6-3-5'> 6.3.5. Impute missing values </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"X_test_2017 = X_test_2017.fillna(method='ffill', axis=1)\nreduce_memory_usage(X_test_2017)\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_test_2018 = X_test_2018.fillna(method='ffill', axis=1)\nreduce_memory_usage(X_test_2018)\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## <a id = '6-4'> 6.4. Make sure test data shape and check for missing values </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{"trusted":true},"cell_type":"code","source":"print('2017 Test Data Shape:', X_test_2017.shape)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print('2018 Test Data Shape:', X_test_2018.shape)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## <a id = '6-5'> 6.5. Generate predictions </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{},"cell_type":"markdown","source":"I tried to generate the predictions with the as-is X_test however, I ran into the memory errors:\n> Unable to allocate array with shape (41498571, 18) and data type float64)\n\nSo I am going to convert some float columns to int data types to occupy less space in the memory."},{"metadata":{"trusted":true,"_kg_hide-output":true},"cell_type":"code","source":"X_test_2017.memory_usage()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-output":true},"cell_type":"code","source":"X_test_2017.dtypes","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# features that datatypes to be converted\nint_features = ['building_age', 'primary_use', \n                'is_weekend',  'wind_compass_direction']\n\nfor feature in int_features:\n    X_test_2017[feature] = X_test_2017[feature].astype('int8')\n    X_test_2018[feature] = X_test_2018[feature].astype('int8')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-output":true},"cell_type":"code","source":"X_test_2017.memory_usage()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-input":true,"_kg_hide-output":true},"cell_type":"code","source":"gc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"If you remember, we took the np.log1p (natural logarithm of meter reading+1) of the meter reading values and trained our model with those values. So, our model learned to predict np.log1p values. \n\nTo align the predictions back to the meter reading values, I am going to use the inverse function of log1p, np.expm1 and generate predictions for 2017 and 2018 seperately due to high memory usage.\n\nnp.expm1 function calculates the  exp(x) - 1 for all elements in the array.\n\nLooking at the training and validation errors of this notebook and previous notebook I expect to see RMSE between 1.2 and 2.5."},{"metadata":{"trusted":true},"cell_type":"code","source":"predictions_2017 = []\n\nfor model in models:\n    if  predictions_2017 == []:\n        predictions_2017 = (np\n                            .expm1(model\n                                   .predict(X_test_2017, \n                                            num_iteration=model.best_iteration)) / len(models))\n    else:\n        predictions_2017 += (np\n                             .expm1(model\n                                    .predict(X_test_2017,\n                                             num_iteration=model.best_iteration)) / len(models))","execution_count":null,"outputs":[]},{"metadata":{"_kg_hide-input":true,"_kg_hide-output":true,"trusted":true},"cell_type":"code","source":"del X_test_2017\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"predictions_2018 = []\n\nfor model in models:\n    if  predictions_2018 == []:\n        predictions_2018 = (np\n                            .expm1(model\n                                   .predict(X_test_2018, \n                                            num_iteration=model.best_iteration)) / len(models))\n    else:\n        predictions_2018 += (np\n                             .expm1(model\n                                    .predict(X_test_2018, \n                                             num_iteration=model.best_iteration)) / len(models))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"del X_test_2018\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# <a id = '7'> 7. Interpret Model Results </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{},"cell_type":"markdown","source":"In the feature engineering part, I selected features based on its pearson coefficients to the log_meter_reading values.  Since final results are generated, I will revisit this picture and observe features with the highest contribution to the model predictions. `plot_importances`  attribute will be used for this, visualizing each of the four model's feature importance individually."},{"metadata":{"trusted":true},"cell_type":"code","source":"for model in models:\n    lgb.plot_importance(model)\n    plt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**In all of the 4 models built, top 15 features for determining the meter reading values:**\n* building_id\n* meter\n* month\n* day of month\n* square_feet\n* season\n* air temperature\n* dew temperature\n* building age\n* site id\n* primary use\n* day of week\n* mean dew tempereature and sea level pressure\n* mean air temperature precip depth 1 hr\n* day of week"},{"metadata":{},"cell_type":"markdown","source":"# <a id = '8'> 8. Submissions & Summary & Conclusions  </a>\n<a href = '#top'> Back to top </a>"},{"metadata":{},"cell_type":"markdown","source":"Thank you for sticking with me until the very end! 😍\n\nIn this notebook, we covered:\n\n6. Evaluate model with test data\n7. Interpret model results\n8. Submissions & summary & conclusions\n\nBefore summarizing the project I will submit predicitons. As I predicted 2017 and 2018 results seperately, I am going to merge them now into one dataframe."},{"metadata":{"trusted":true},"cell_type":"code","source":"# to fetch row_ids\nsample_submission = pd.read_csv(\"/kaggle/input/ashrae-energy-prediction/sample_submission.csv\")\nrow_ids = sample_submission.row_id\n\ndel sample_submission\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# make sure of the shape of predictions\npredictions_2017.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"predictions_2018.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# split row_id's with the indexes of 2017 and 2018 predictions\nrow_ids_2017 = row_ids[:predictions_2017.shape[0]]\nrow_ids_2018 = row_ids[predictions_2018.shape[0]:]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission_2017 = pd.DataFrame({\"row_id\": row_ids_2017, \n                                \"meter_reading\": np.clip(predictions_2017, 0, a_max=None)})\n\nsubmission_2018 = pd.DataFrame({\"row_id\": row_ids_2018, \n                                \"meter_reading\": np.clip(predictions_2018, 0, a_max=None)})","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission = pd.concat([submission_2017,\n                        submission_2018])\n\ndel submission_2017, submission_2018\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission.to_csv(\"submission.csv\", index=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"del models\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"We've come to an end in this long and comprehensive project, the highlights from the 3 notebooks and the machine learning workflow are: \n\n* Exploratory data analysis\n* Feature engineering & selection\n* Building an efficient model with model selection, tuning and cross-validation\n\nIn the [first notebook](https://www.kaggle.com/cereniyim/save-the-energy-for-the-future-1-detailed-eda), first I understand analyzed what affects most a building's energy consumption in electricity, chilled water, steam and hot water. The area, purpose and age are the most important factors for a building to determine its energy consumption. The air temperature, humidity, sea level pressure and how cloudy the sky is at the current time are the most important weather factors for building's energy consumption.\n\nIn the [second notebook](om/cereniyim/save-the-energy-for-the-future-2-fe-lightgbm/notebook), feature engineering showed its importance: with the newly generated features as part of this project had also high impact in determining the meter_reading values.\n\nI moved forward with the light gbm model which seems as efficient solution when dealing with 20 or 40 million rows after exploring some other notebooks in this competition. Based on the 30 features selected and 1 year time-series data available (nearly 20 million rows) I predicted the next 2 years electricity, chilled water, hot water and steam consumption of each building (nearly 40 million rows). Even though I did not climb high in the competition ladder I learned a lot on the way which is more important for me.\n\nDuring this project, I find a chance to explore new libraries such as pandas profiling, hyperopt and light gbm and learned to deal with large dataset as well as the RAM. Most importantly, I applied the machine learning workflow to a supervised regression problem for the first time, introduced by Will Koehrsen [in this article](https://towardsdatascience.com/a-complete-machine-learning-walk-through-in-python-part-one-c62152f39420).\n\nThank you everyone for reading, upvoting and commenting on my kernels. I hope that I contributed to make energy-efficient buildings with this project and inspire other people to join this cause with me. "},{"metadata":{},"cell_type":"markdown","source":"![image](https://i.pinimg.com/736x/7c/6e/1b/7c6e1b412a365376894828a348955372.jpg)\nImage source: [pinterest](https://www.pinterest.com/pin/576179346069675606/) "}],"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":4,"nbformat_minor":1}