import numpy as np
#import matplotlib.pyplot as plt
import pandas as pd
import math

def debug(y):
    print(y)
    exit()

data = pd.read_csv('/content/drive/MyDrive/train.csv') #add file path //C:
data_test = pd.read_csv('/content/drive/MyDrive/test.csv') #add file path
rawData = np.asarray(data_test)

x_test = []

def convert(x):
    if (x == 'S'): return 1
    elif (x == 'Q'): return 2
    return 3

cnt = 0
val = 0
for i in range (0,len(rawData)):
    if (math.isnan(rawData[i][4])): continue
    else:
        cnt += rawData[i][4]
        val += 1
med = val/cnt

for i in range(0,len(rawData)):
    x_test.append([])
    x_test[i].append(rawData[i][1])

    if (rawData[i][3] == 'male'): x_test[i].append(1)
    else: x_test[i].append(0)
    #x_test[i].append(rawData[i][3])

    if (math.isnan(rawData[i][4])): x_test[i].append(med)
    else: x_test[i].append(rawData[i][4])

    x_test[i].append(rawData[i][5])
    x_test[i].append(rawData[i][6])
    x_test[i].append(rawData[i][8])

    x_test[i].append(convert(rawData[i][10]))

x_test = np.asarray(x_test)

x_train = []
y_train = []

survive = np.asarray(data.Survived);
for i in range (0,891):
    x_train.append([])
    y_train.append(np.array([survive[i]]))

def Add_To_X(a):
    for i in range(0,len(a)): x_train[i].append(a[i])

pclass = np.asarray(data.Pclass)
Add_To_X(pclass)

gender = np.asarray(data.Sex)
for i in range(0,len(gender)):
    if gender[i] == 'male': x_train[i].append(1)
    else: x_train[i].append(0)

age = np.asarray(data.Age)
Add_To_X(age)

SibSp = np.asarray(data.SibSp)
Add_To_X(SibSp)

Parch = np.asarray(data.Parch)
Add_To_X(Parch)

fare = np.asarray(data.Fare)
Add_To_X(fare)

embark = np.asarray(data.Embarked)
for i in range (0,891):
    if (embark[i] == 'S'): x_train[i].append(1)
    elif (embark[i] == 'Q'): x_train[i].append(2)
    else: x_train[i].append(3)

cnt1 = 0
val1 = 0
for i in range (0,len(x_train)):
    if (math.isnan(x_train[i][2])): continue
    else:
        cnt1 += x_train[i][2]
        val1 += 1
med1 = val1/cnt1

for i in range (0,len(x_train)):
    if (math.isnan(x_train[i][2])): x_train[i][2] = med1

x_train = np.asarray(x_train)
y_train = np.asarray(y_train)

import numpy as np
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
import matplotlib.pyplot as plt
import logging

print(x_train.shape)
print(x_test.shape)
exit()

def my_leaky_relu(x):
    return tf.nn.leaky_relu(x, alpha=0.003)

model = Sequential([
    Dense(units = 2048*3,activation = 'relu'),
    Dense(units = 2048*3,activation = 'relu'),
    Dense(units = 1024*3,activation = 'relu'),
    Dense(units = 1024*3,activation = 'relu'),
    Dense(units = 512*3,activation = 'relu'),
    Dense(units = 1,activation = 'sigmoid')
])

model.compile(
    loss=tf.keras.losses.BinaryCrossentropy(),
    optimizer=tf.keras.optimizers.Adam(0.001),
)
model.fit(x_train,y_train,epochs = 1500)

prediction = model.predict(x_test)
for i in range (0,len(prediction)):
    if (prediction[i] >= 0.5): prediction[i] = 1
    else: prediction[i] = 0

#print(prediction[0][0])

total = []
for i in range (0,len(prediction)):
    now = []
    now.append(892 + i)
    if (prediction[i][0] == 0): now.append(0)
    else: now.append(1)
    total.append(now)

#print(total[0])

df = pd.DataFrame(total)
df.to_csv('ans.csv')