controller.py

import pandas as pd
import numpy as np
from sklearn.ensemble import RandomForestClassifier #rforest

from entity import Entity
from service import Service

class Controller:

# self.entity => entity is an attribute of Controller instance. 
    def __init__(self):
        self.entity = Entity()
        self.service = Service()

    def modeling(self, train, test):
        service = self.service
        this = self.preprocessing(train, test)
        this.label = service.create_label(this)
        this.train = service.create_train(this)
        print(f'>> Train variables : {this.train.columns}')
        print(f'>> Test variables : {this.train.columns}')
        return this

    def preprocessing(self, train, test):
        service = self.service
        this = self.entity
        this.train = service.new_model(train) # payload
        this.test = service.new_model(test) # payload
        this.id = this.test['PassengerId'] # machine 이에게는 이것이 question 이 됩니다.
        print(f'Train variables before processing : {this.train.columns}')
        print(f'Test variables before processing: {this.test.columns}')
        this = service.drop_feature(this, 'Cabin')
        this = service.drop_feature(this, 'Ticket')
        print(f'Variables after drop : {this.train.columns}')
        this = service.embarked_norminal(this)
        print(f'Embarked ports after processing: {this.train.head()}')
        this = service.title_norminal(this)
        print(f'Titles after processing: {this.train.head()}')
        # We extracted title out of name => no need for name anymore
    
        this = service.drop_feature(this, 'Name')
        this = service.drop_feature(this, 'PassengerId')
        this = service.age_ordinal(this)
        print(f'Age after processing: {this.train.head()}')
        this = service.drop_feature(this, 'SibSp')
        this = service.sex_norminal(this)
        print(f'Sex after processing: {this.train.head()}')
        this = service.fareBand_nominal(this)
        print(f'Fare after processing: {this.train.head()}')
        this = service.drop_feature(this, 'Fare')
        print(f'#########  Result of processing for TRAIN  ###############')
        print(f'{this.train.head()}')
        print(f'######### Result of processing for TEST ###############')
        print(f'{this.test.head()}')
        print(f'######## check train na ##########')
        print(f'{this.train.isnull().sum()}')
        print(f'######## check test na ##########')
        print(f'{this.test.isnull().sum()}')
        return this
        

    def learning(self, train, test):
        service = self.service
        this = self.modeling(train, test)
        print('&&&&&&&&&&&&&&&&& Learning Results  &&&&&&&&&&&&&&&&')
        print(f'Dtree: {service.accuracy_by_dtree(this)}')
        print(f'Random Forest: {service.accuracy_by_rforest(this)}')
        print(f'NB: {service.accuracy_by_nb(this)}')
        print(f'KNN: {service.accuracy_by_knn(this)}')
        print(f'SVM: {service.accuracy_by_svm(this)}')

    def submit(self,train, test):
        this = self.modeling(train,test)
        clf = RandomForestClassifier()
        clf.fit(this.train, this.label)
        prediction = clf.predict(this.test)
        pd.DataFrame(
        {'PassengerId' : this.id, 'Survived': prediction}).to_csv(this.context + 'submission.csv', index =False)


if __name__ == '__main__':
    ctrl = Controller()
    ctrl.learning('train.csv', 'test.csv')
    # ctrl.submit('train.csv', 'test.csv')