kusal1990

challenge_train_preds = []
challenge_dev_preds = []
challenge_test_preds = []

for i in range(len(df_challenge_train)):
    if '(1)' in df_challenge_train['only_answers'].iloc[i]:
        challenge_train_preds.append(random.choice(num_list))
    else:
        if '(E)' in df_challenge_train['only_answers'].iloc[i]:
            challenge_train_preds.append(random.choice(char_list_2))

kusal1990

easy_train_preds = []
easy_dev_preds = []
easy_test_preds = []

for i in range(len(df_easy_train)):
    if '(1)' in df_easy_train['only_answers'].iloc[i]:
        easy_train_preds.append(random.choice(num_list))
    else:
        if '(E)' in df_easy_train['only_answers'].iloc[i]:
            easy_train_preds.append(random.choice(char_list_2))

kusal1990

def get_context(question, options):
    '''
    This function will return a context after joining a question
    and its options separately. 
    For example: question + option_1
                 question + option_2
                 question + option_3
                 question + option_4
   '''
    scores = []

kusal1990

def get_context_for_each_candidate(question,options_list):
    """ Searches the user query and finds the best matches using elasticsearch."""
    #query = input("Enter query: ")
    """This function will return a context after joining a question
    and its options separately. 
    For example: question + option_1
                 question + option_2
                 question + option_3
                 question + option_4"""
   

kusal1990

def data_generator(a):
    '''
    This function will accept a data point and returns a list of options
    converts the only_options list of options.

    Output will be as follows:

    options = [option_1, option_2, option_3, option_4, option_5]                   

    Note: If there's no option E then it will be written as 'None of the above'.

kusal1990

from elasticsearch import Elasticsearch
from elasticsearch.helpers import parallel_bulk

import pandas as pd
import numpy as np
import re

es = Elasticsearch([{'host':'localhost', 'port':9200}])

corpus2 = open('drive/MyDrive/ARC/ARC_Corpus.txt', 'r', encoding='utf-8')

kusal1990

def extract_answers(question):
    '''
    This function will take a value in question column then, will split the string 
    at (A) or (1). After splitting we will take only first part which is the question.
    '''
    if '(A)' in question:
        x = '(A) ' + question.split('(A)')[1]
    elif '(1)' in question:
        x = '(1) ' + question.split('(1)')[1]

kusal1990

def extract_question(question):
    '''
    This function will take a value in question column then, will split the string 
    at (A) or (1). After splitting we will take only first part which is the question.
    '''
    if '(A)' in question:
        x = question.split('(A)')[0]
    elif '(1)' in question:
        x = question.split('(1)')[0]

kusal1990

def Year_graph(type_):
    print(type_+" dataset Analysis ...............")
    for i,v in enumerate(["Train","Test","Dev"],start = 1):
        plt.figure(figsize=(9,6))
        examName_dict = dict(eval(v+"_multiple_main_"+ type_ +"[\"year\"].value_counts()"))
        plt.title(v+"_ExamName vs No_of_qustions_of_particular_catagory("+type_+")")
        plt.barh([i for i in examName_dict.keys()],[i for i in examName_dict.values()])
        plt.xlabel("No_of_qustions_of_particular_catagory")
        plt.show()

kusal1990

def Exam_graph(type_):
    print(type_+" dataset Analysis ...............")
    for i,v in enumerate(["Train","Test","Dev"],start = 1):
        plt.figure(figsize=(9,6))
        examName_dict = dict(eval(v+"_multiple_main_"+ type_ +"[\"examName\"].value_counts()"))
        plt.title(v+"_ExamName vs No_of_qustions_of_particular_catagory("+type_+")")
        plt.barh([i for i in examName_dict.keys()],[i for i in examName_dict.values()])
        plt.xlabel("No_of_qustions_of_particular_catagory")
        plt.show()