Z4HRA-S

from datetime import datetime
import cProfile
import time
import logging
import multiprocessing as mp
import psutil

class CumulativeProfile:
    def __init__(self, func, v=0):
        self.__name__ = func.__name__

Z4HRA-S

from models import SaleRecords
from sqlalchemy import func


def sold_product_percent(session) -> dict:
    sold_product = session.query(func.sum(SaleRecords.count), SaleRecords.product_name
                          ).group_by(SaleRecords.product_name).all()
    sum_of_products = sum(count for count, name in sold_product)
    percents = {item[1]: round((item[0] / sum_of_products) * 100, 2)
                for item in sold_product}

Z4HRA-S

def test_decoupled_sold_product_percent(self):
    # Arrange
    session = mock.Mock()
    session.query.return_value = session
    session.filter.return_value = session
    session.where.return_value = session
    session.group_by.return_value = session
    session.all.return_value = [(76, 'cheese_cake'),
                                (67, 'chocolate_cake'),
                                (70, 'coffee'),

Z4HRA-S

class TestMain(unittest.TestCase):

    def test_sold_product_percent(self):
        # Arrange
        session = mock.Mock()
        session.query().group_by().all.return_value = [(76, 'cheese_cake'),
                                                       (67, 'chocolate_cake'),
                                                       (70, 'coffee'),
                                                       (17, 'hot_dog'),
                                                       (70, 'tea')]

Z4HRA-S

Thesis Summary

The task of relation extraction focuses on finding relational information between entities in a text. The input text can be one sentence or a document. Relational facts are highly semantic information that needs reasoning skills to discover. Therefore, the model must be capable of processing the text's semantic concepts. In this thesis, we explored the potential of the Caps-net in extracting high-level semantic information, capturing the relational information from the text, and the ability to overcome the imbalance problem of the data set. Our dataset was a complex data set with a severe imbalance and a relatively large number of potential entity-pair per document. The large number of entity pairs along with the complexity needed for solving this task, make the whole proposed model computationally expensive. In addition to trying to improve the model's performance on the dataset, we tried to reduce the memory and time complexity of the proposed model. In this work, we studied the Caps-Net [