ML Kaggle tatanic

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from sklearn.preprocessing import Imputer
from sklearn.preprocessing import LabelEncoder, OneHotEncoder
from sklearn.linear_model import LinearRegression

# Importing the dataset
dataset = pd.read_csv('train.csv')
df = pd.DataFrame(dataset)

# Replace nan values in Embarked with most frequent element
df.Embarked = df.Embarked.fillna(df['Embarked'].value_counts().idxmax())
X = df.iloc[:, [2, 4, 5, 6, 7, 9 ,11]].values
y = df.iloc[:, 1].values
# check for nan
#df.Embarked.isnull().values.any()


# Replace nan values in Age with mean of values
imputer = Imputer(missing_values = 'NaN', strategy = 'mean', axis = 0)
imputer = imputer.fit(X[:, 2:3])
X[:, 2:3] = imputer.transform(X[:, 2:3])

# Encoding categorical data
labelencoder_X = LabelEncoder()
X[:, 1] = labelencoder_X.fit_transform(X[:, 1])
labelencoder_X_2 = LabelEncoder()
X[:, 6] = labelencoder_X_2.fit_transform(X[:, 6])

onehotencoder = OneHotEncoder(categorical_features = [1])
X = onehotencoder.fit_transform(X).toarray()
# Avoiding the dummy variable trap
X = X[:, 1:]
onehotencoder_2 = OneHotEncoder(categorical_features = [6])
X = onehotencoder_2.fit_transform(X).toarray()
# Avoiding the dummy variable trap
X = X[:, 1:]

# predicting with linear regression
regressor = LinearRegression()
regressor.fit(X,y)

# test part
dataset_test = pd.read_csv('test.csv')
df_test = pd.DataFrame(dataset_test)

X_test_id = df_test.iloc[:, [0]].values

# Replace nan values in Embarked with most frequent element
df_test.Embarked = df_test.Embarked.fillna(df_test['Embarked'].value_counts().idxmax())
X_test = df_test.iloc[:, [1, 3, 4, 5, 6, 8 ,10]].values
# check for nan
#df.Embarked.isnull().values.any()
for x in X_test:
    print(x)

# Replace nan values in Age with mean of values
imputer_test = Imputer(missing_values = 'NaN', strategy = 'mean', axis = 0)
imputer_test = imputer_test.fit(X_test[:, 2:3])
X_test[:, 2:3] = imputer_test.transform(X_test[:, 2:3])
imputer_test_2 = Imputer(missing_values = 'NaN', strategy = 'mean', axis = 0)
imputer_test_2 = imputer_test_2.fit(X_test[:, 5:6])
X_test[:, 5:6] = imputer_test_2.transform(X_test[:, 5:6])

# Encoding categorical data
labelencoder_X_test = LabelEncoder()
X_test[:, 1] = labelencoder_X_test.fit_transform(X_test[:, 1])
labelencoder_X_2_test = LabelEncoder()
X_test[:, 6] = labelencoder_X_2_test.fit_transform(X_test[:, 6])

onehotencoder_test = OneHotEncoder(categorical_features = [1])
X_test = onehotencoder_test.fit_transform(X_test).toarray()
# Avoiding the dummy variable trap
X_test = X_test[:, 1:]
onehotencoder_2_test = OneHotEncoder(categorical_features = [6])
X_test = onehotencoder_2_test.fit_transform(X_test).toarray()
# Avoiding the dummy variable trap
X_test = X_test[:, 1:]

y_pred = regressor.predict(X_test)

y_pred = [round(x) for x in y_pred]

y_pred_int = [round(x) for x in y_pred]

y_pred_int = [int(i) for i in y_pred_int]

with open('pred.csv','wb') as file:
    for i in range(len(y_pred_int)):
        file.write(X_test_id[i] + ',' + y_pred_int[i])
        file.write('\n')
file.close()