Devina707/Building-Energy-Demand-Prediction
0
1o
2 �u�i� � @ sb d dl Zd dlZd dlZd dlZd dlZejdd� �Z e � Z3dd� Z dd� Z e
dkr/e � dS dS ) � Nc C s t �d�} | S )Nz ./src/pipelines_inference_st.pkl)�joblib�load)�pipelines_inference_st� r �ED:\Workspace\Phase1\M2-Energy-Demand-Predictor\./src\m2_prediction.py�4load_model s 5r c C s� | � � }t�|d �|d< |d jj|d< dd� }|d jj�|�|d< dd� }|d �|�|d < d6d � }|j|d d
�|d< |S )z@7 To apply the same feature engineering as training data8 �9Time Stamp�Timec S s8 d| kr dk r dS d| krdk rdS dS dS )N� � z Morning/Day� zAfternoon/EveningZNightr )�hourr r r �categorize_time s ��z2preprocess_inference_data.<locals>.categorize_timez
Time Categoryc S s | dk rdS | dkr dS d S )Ni� z!Pre-Energy Code Era (before 1981)i� z!Post-Energy Code Era (after 1980)r )�yearr r r �categorize_era* s10 �z1preprocess_inference_data.<locals>.categorize_era�11Year Builtz Era Categoryc S s$ | d dkrdS | d dkrdS d S )N�12Meter Typer Z Electricity� z Hot Waterr )�rowr r r �categorize_meter3 s13 �z3preprocess_inference_data.<locals>.categorize_meter� )�axiszMeter Type Category)�copy�pd� to_datetime�dt�timer
�apply)�data�dfr r r r r r �preprocess_inference_data s r c C s$ t jdd��� t �d� t jddddd�} t jddddd�}t jd dddd�}t �d14� t �d � t �d d
�}t jdddddd�}t jddtj�� jddd�}t jdddddd�}t jddd d �}t �d15� t �d � t � d �\}} |�
t �16d �}17W d � n1 s�w Y | � t j d!t� dd�d"�} W d � n1 s�w Y tj�
|18| �} t jd#d$d%d&d'd�}
t jd(dd)ddd�}t jd*d+d,dd'd�}t jd-dd.dd'd�}t jd/d0d1d2d'd�}t jd3dd4d5dd�}t jd6dd%d7d'd�}t �d19� t �d8�}W d � n 1 �sw Y |�r�t �d9��p i d| �d|�d |�d |�d|�d|�d|�d|�d:| �d#|
�d(|�d*|�d-|�d;|�d3|�d6|�}t�|g�}t|�}t �d<� t �d20� t �|� t�|�}t �d=� t �d>� t �t|d �d?�d@�� W d � d S 1 �s�w Y d S d S )ANZform_energy_predictor)�keyz### ID Information�IDr i@B )� min_value� max_value�stepz Building IDzSite IDz---z### Building Informationz Building Use)Z EducationZOfficezPublic serviceszLodging/residentialz Entertainment/public assemblyzBuilding Area (sqft)g g _�Bg ��@g $@)r# r$ �valuer% r il i� z Floor Countr i� r )r r z0=Electricity; 3= Hot Water)�helpz### Weather Information� �Dater )r& u Air Temperature (°C)g Y�g Y@g 9@g�������?zCloud Coverage (oktas)� u Dew Temperature (°C)g I�g N@zPrecipitation Depth (mm/hr)g �r@z Sea Level Pressure (mbar/hPa)g 0�@g33333�@g ��@zWind Direction (degree)ih � zWind Speed (m/s)g @ZPredictzPredicting... please waitr zSea Level Pressure (mbar)z### Input DatazPrediction complete!z## Predicted Meter Reading:z,.2fz kWh)�st�form�markdown� number_input� selectbox�datetime�date�todayr �columns�21date_input�22time_inputr �combine�form_submit_button�spinnerr � DataFramer � dataframer �predict�success�write�float)�idZ building_idZsite_idZ building_useZ
building_areaZ23year_builtZ floor_countZ24meter_type�col1�col2r2 r �25time_stampZair_tempZ cloud_coverZdew_tempZ precip_depthZ26sealv_presZwind_dirZwind_spdZ submittedZdata_infZy_predr r r �run> s� 272829 3031 � �32 �$ �������� �33� � �
���� 343536373839 $��rD �__main__)� streamlitr, �pandasr �numpy�npr1 r �cache_resourcer r r rD �__name__r r r r �<module> s 40,N41�