def classify_new_sounds(folder_of_test_sounds, target_class):
new_descriptors = {target_class: load_description(folder_of_test_sounds)}
new_normalized_features, new_y_class, new_features_names = preprocessDescriptors(new_descriptors)
new_y_class = np.array([target_class]*len(new_y_class))
descriptors = loadDescriptors(maximum='Inf', reverbs=True)
normalized_features, yClass, features_names = preprocessDescriptors(descriptors)
clf = trainSVM(normalized_features, yClass, call=True)
F1 = F1Score(new_normalized_features, new_y_class, clf)
return F1
def classify_new_sounds(folder_of_test_sounds, target_class):
new_descriptors = {target_class: load_description(folder_of_test_sounds)}
new_normalized_features, new_y_class, new_features_names = preprocessDescriptors(
new_descriptors)
new_y_class = np.array([target_class] * len(new_y_class))
descriptors = loadDescriptors(maximum='Inf', reverbs=True)
normalized_features, yClass, features_names = preprocessDescriptors(
descriptors)
clf = trainSVM(normalized_features, yClass, call=True)
F1 = F1Score(new_normalized_features, new_y_class, clf)
return F1
def run_stats_analysis(maximum='Inf', reverbs=True):
Descriptors = loadDescriptors(maximum, reverbs)
normalized_features, yClass, features_names = preprocessDescriptors(Descriptors)
every_f1_test, mMean, mVar = test_repe(100, normalized_features, yClass)
instrument_count = count_instruments(yClass)
mean = mean_per_instrument(every_f1_test)
stand = std_per_instrument(every_f1_test)
return mean, stand, instrument_count
def run_stats_analysis(maximum='Inf', reverbs=True):
Descriptors = loadDescriptors(maximum, reverbs)
normalized_features, yClass, features_names = preprocessDescriptors(
Descriptors)
every_f1_test, mMean, mVar = test_repe(100, normalized_features, yClass)
instrument_count = count_instruments(yClass)
mean = mean_per_instrument(every_f1_test)
stand = std_per_instrument(every_f1_test)
return mean, stand, instrument_count
# -*- coding: utf-8 -*-
from sklearn import svm
import numpy as np
import matplotlib.pyplot as plt
from utils.FScore import F1Score
from Identification.LoadDescriptors import loadAllDescriptors
from Identification.PreprocessingDescriptors import preprocessDescriptors
from Identification.TrainCvTest import separateDatabases
Descriptors = loadAllDescriptors(reverbs=True)
normalized_features, yClass, features_names = preprocessDescriptors(Descriptors)
del Descriptors # Ya no lo voy a utilizar
normalizedTrain, yTrain, normalizedCV, yCV, normalizedTest, yTest = separateDatabases(normalized_features, yClass)
def test_data_size(training_features, training_classes, test_features, test_classes):
index = np.arange(0, len(training_classes))
np.random.shuffle(index)
test_size = np.linspace(0.1, 1, 50) * len(index)
test_size = [int(i) for i in test_size]
f_train = []
f_cv = []
clf = svm.SVC(C=1.833, gamma=0.1366, cache_size=1000)
for iii in test_size:
clf.fit(training_features[index[0:iii]], training_classes[index[0:iii]])
f_train = np.append(f_train, np.mean(F1Score(training_features[index[0:iii]],