from matplotlib import pyplot as plt
from sklearn.neighbors import KNeighborsClassifier as nbh
from sklearn import metrics
import pandas as pd
from sklearn import preprocessing
import numpy as np
from sklearn.metrics import accuracy_score


def make_plot(ratios, accuracies, title):
    plt.figure()
    plt.plot(ratios, [acc[0] for acc in accuracies], label='test data')
    plt.plot(ratios, [acc[1] for acc in accuracies], label='train data')
    plt.xlabel('training data size')
    plt.ylabel('accuracy')
    plt.title(f'{title}\naccuracy(training data size)')
    plt.legend()
    plt.savefig(f'{title}.png')


def accuracy(feat, train, tr_size):
    tr_size = 1-tr_size
    neigh = nbh(n_neighbors=3, n_jobs=1)
    neigh.fit(feat, train)
    print(neigh.predict(i for i in feat))


def tic_tac_toe():
    features, targets = [], []
    with open("Tic_tac_toe.txt") as inp:
        for line in inp:
            features.append(line.split(',')[0:9])
            targets.append(line.split(',')[9].strip())
    le = preprocessing.LabelEncoder()
    features_encoded = [le.fit_transform(sample) for sample in features]
    targets_encoded = le.fit_transform(targets)
    ratios = np.linspace(0.01, 0.9, 100)
    accuracies = [accuracy(features_encoded, targets_encoded, ratio) for ratio in ratios]
    make_plot(ratios, accuracies, 'tic-tac-toe')


tic_tac_toe()