main.py

import json
import pandas as pd
import matplotlib.pyplot as plt

with open("responses.json") as file:
    data = json.load(file)

list_data = [i for i in data.values()]


# Each JSON entry (i.e. person) is a dictionary
# Calculate compatibility score for two people
def score(person1, person2):
    # Assign each question an arbitrary weight of 10 for now
    weights = [10, 10, 10, 10, 10, 10, 10, 10, 10, 10]
    # Find the question similarity score
    question_score = 0
    for i in range(len(person1["answers"])):
        if person1["answers"] == person2["answers"]:
            question_score += weights[i] * 1
        else:
            question_score += weights[i] * 0.20

    # Year compatibility
    year_score = 1 / (abs(person1["year"] - person2["year"]) + 1)

    # Gender preference compatiblity
    gender_score = 0
    if person1["gender"] in person2["gender_preferences"] and person2["gender"] in person1["gender_preferences"]:
        gender_score = 1

    # Similarity preference compatibility
    f_score = (person1["f"] + person2["f"]) / 2

    return gender_score * year_score * abs(1 / (question_score / 100 - f_score + 1))


print(score(data["2"], data["3"]))

pair = []
# If want to plot non-zero pairs
for person1 in data:
    for person2 in data:
        if person1 < person2:
            scr = score(data[person1], data[person2])
            if scr != 0:
                pair.append(scr)


# If want to plot all pairs
# for person1 in data:
#     for person2 in data:
#         if person1 < person2:
#             pair.append(score(data[person1], data[person2]))

# print(len(pair))

# Create dataframe of pair list
df = pd.DataFrame(pair, columns=['Score'])

# matplotlib histogram
plt.hist(df['Score'], color='blue', edgecolor='black', bins=int(180 / 5))

# Add labels
plt.title('Histogram of Pair Scores')
plt.xlabel('Scores')
plt.ylabel('Count')

plt.show()

def pair():
    rating_rankings = {}
    # Calculate all possible ratings
    # Add to the list if not 0
    size = len(list_data)
    for i in range(size):
        rating_rankings[i] = []

    for i in range(size):
        for k in range(i, size): 
            scr = score(list_data[i], list_data[k])
            if scr > 0:
                rating_rankings[i].append((k, scr))
                rating_rankings[k].append((i, scr))

    # Sort all the preferences
    for i in range(size):
        rating_rankings[i].sort(key=lambda x: x[1], reverse=True) 

    # print(rating_rankings)

    # # Perform the matching
    # pairs = {}
    # for i in range(size):
    #     for k in range(i, size): 
    #         scr = score(list_data[i], list_data[k])
    #         # Case 1: i'th person and k'th person are not in dictionary (i.e. unpaired)  
    #         if i not in pairs and k not in pairs: 
    #             pairs[i] = [scr, k]
    #             pairs[k] = [scr, i]
    #         # Case 2: k is unpaired by i already has a pair
    #         if k not in pairs and scr > pairs[i][1]:
    #             pairs[i] = [scr, k]
    #             pairs[k] = [scr, i]
    #         # Case 3: Both i and k already have pairs
    #         if scr > pairs[i][1] and scr > pairs[k][1]:
    #             pairs[i] = [scr, k]
    #             pairs[k] = [scr, i]


# pair()