import pandas as pd
try:
    from StringIO import StringIO
except ImportError:
    from io import StringIO
from io import StringIO
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split
from sklearn import metrics
from sklearn.tree import export_graphviz
import joblib
from IPython.display import Image
import pydotplus
import os
os.environ["PATH"] += os.pathsep + r'C:\Program Files (x86)\Graphviz2.38\bin'

def main():
    dot_data = StringIO()
    col_names = ['age', 'sex', 'fat', 'fiber', 'spicy', 'number']
    #import danych
    model_tree = pd.read_csv("Nowy.csv", header=None, names=col_names)
    model_tree.head()
    #seperacja danych cechy
    feature_cols = ['age', 'sex', 'fat', 'fiber', 'spicy']
    X = model_tree[feature_cols]
    #separacja danych etykieta
    y = model_tree.number


    #podział danych na zestaw treningowy i testowy; 70% trening 30% test
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3,
                                                        random_state=1)

    ###############################
    # stworzenie -obiektu- drzewa Decision Tree classifer
    clf = DecisionTreeClassifier()

    # drzewo treningowe
    clf = clf.fit(X_train, y_train)

    #generowanie wyników dla zbioru testowego
    y_pred = clf.predict(X_test)
    #print(y_pred)

    #Akuratność dla modelu danych
    print("Accuracy:", metrics.accuracy_score(y_test, y_pred))

    dot_data = StringIO()
    #tworzenie graficznego drzewa
    export_graphviz(clf, out_file=dot_data,
                    filled=True, rounded=True,
                    special_characters=True, feature_names=feature_cols,
                    class_names=['1', '2', '3', '4', '5', '6', '7'])
    graph = pydotplus.graph_from_dot_data(dot_data.getvalue())
    graph.write_png('polecanie_1.png')
    Image(graph.create_png())

    # zapisanie modelu danych do pliku
    file_name = 'final_model.sav'
    joblib.dump(clf, file_name)

    # ************************************************************
    # stworzenie -obiektu- drzewa Decision Tree classifer z kryterium entropii
    clf = DecisionTreeClassifier(criterion="entropy")

    #drzewo testowe-z entripią
    clf = clf.fit(X_train, y_train)

    #generowanie wyników dla zbioru testowego
    y_pred = clf.predict(X_test)

    #Akuratność dla modelu danych z warunkiem entropii
    print("Accuracy:", metrics.accuracy_score(y_test, y_pred))

    #stowrzenie graficznego drzewa z warunkiem entropii
    dot_data = StringIO()
    export_graphviz(clf, out_file=dot_data,
                    filled=True, rounded=True,
                    special_characters=True, feature_names=feature_cols,
                    class_names=['1', '2', '3', '4', '5', '6', '7'])
    graph = pydotplus.graph_from_dot_data(dot_data.getvalue())
    graph.write_png('polecanie_2_entropia.png')
    Image(graph.create_png())



    return

main()