fitness.py

"""
This file is used by main and gives the fitness of a particular set of features.
"""
import os
import time
import xml.etree.ElementTree as ET
import keras
import matplotlib.pyplot as plt
from matplotlib import offsetbox
import matplotlib.patheffects as PathEffects
import seaborn as sns
import numpy as np
import tensorflow as tf
from keras.layers import Dense, Dropout
from keras.models import Sequential
from sklearn.ensemble import GradientBoostingClassifier
from sklearn.metrics import accuracy_score, precision_score
from sklearn.preprocessing import StandardScaler
from sklearn.manifold import TSNE
from sklearn.model_selection import StratifiedKFold
from deferred import dataload_hex
from sklearn import datasets
from mpl_toolkits.mplot3d import Axes3D
#from XGBoost import run_XGBoost

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
print('TensorFlow version: {0}'.format(tf.__version__))

# The data has to be same that was used during the evolution of the model. It is used during the standardization phase.
X1 = ET.parse(
    r"jSymbolic2\features\extracted_feature_led_zep124.xml").getroot()
X2 = ET.parse(
    r"jSymbolic2\features\extracted_feature_random_feat.xml").getroot()
X3 = ET.parse(
    r"jSymbolic2\features\extracted_feature_mid_test.xml").getroot()

# Path to the chosen model. Don't forget to change it when you want to use a different model.
model_path = "models/dense_xml_ledzep_random_1.000.h5"


def parse(X, y):
    out = []
    for song in X[1:]:  # removes the first
        features = []
        for feature in song[1:]:  # removes the first
            features.append(float(feature[1].text.replace(',', '.')))  # commas in XML files have to be turned into dot
        out.append(features)
    return out, [y for k in range(len(X[1:]))]


def get_data(X1, X2, X3=None, X4=None):
    xt1, yt1 = parse(X1, 0)
    xt2, yt2 = parse(X2, 1)
    xt3, yt3, xt4, yt4 = None, None, None, None
    if X3 is not None:
        xt3, yt3 = parse(X3, 2)
    if X4 is not None:
        xt4, yt4 = parse(X4, 3)

    X_tot = np.concatenate(tuple((k for k in (xt1, xt2, xt3, xt4) if k is not None)))
    y_tot = np.concatenate(tuple((k for k in (yt1, yt2, yt3, yt4) if k is not None)))
    return X_tot, y_tot


# Standardization with regard to the original data is needed in order to have the same normalization
X_tot, y_tot = get_data(X1, X2)
scaler = StandardScaler().fit(X_tot)

# Calls the model
model = keras.models.load_model(model_path)


def fitness(features):
    # Standardizes the input
    X_test = scaler.transform(features)

    hist = model.predict(X_test, verbose=0)
    return [hist[k][0] for k in range(len(hist))]

'''verif = [k[0]>0.5 for k in hist]

print("\nScore :", verif.count(True)/len(verif))

for k in range(len(verif)):
    if not verif[k]:
        print(X3[k][0].text)
        print("Confidence :", hist[k][0]*100, "%")'''