import torch
import torch.nn as nn
import auto_training as at
import func
import numpy as np
import csv
import pandas as pd
sem = input('Enter the semester: ')
# parameter
INF = 123.0
K = 10
# prepare data and model
stds = func.findAllStudent()
allCos = func.findAllCos()
score = func.findGrades(stds, allCos)
score = np.float32(np.nan_to_num(score))
score = torch.FloatTensor(score)
currentCos = func.findCurrentCos(sem)
AutoEncoder = at.AutoEncoder
autoencoder = torch.load('net2.pkl')
# start encode
print('Encoding')
encoded, _ = autoencoder(score)
encoded = encoded.detach().numpy()
# get similarity
print('Getting Similarity')
s_len = len(encoded)
import func
import json
sem = '108-1'
current_cos = func.findCurrentCos(sem) # current cos name
all_cos = func.findAllCos() # all cos name
all_student_id = func.findAllStudent() # all student id
## set up id pair in graph
## get student_id and it's passed cos name
## translate it to graph id
id_pair = dict() # graph id pair
max_id = 0
id_cos_pair = func.findIdCos()
res_pair = dict()
for std_id, coses in id_cos_pair.items():
if std_id not in id_pair:
id_pair[std_id] = max_id
max_id += 1
cos = []
for i in coses:
if i not in id_pair:
id_pair[i] = max_id
max_id += 1
cos.append(str(id_pair[i]))
graph_id = str(id_pair[std_id])
res_pair[graph_id] = cos
## output json file
with open('input.json', 'w') as f:
f.write(json.dumps(res_pair))