def loadDataSet(dataSet='ORL', splitNum=3):
if dataSet == 'ORL':
return loadImg(splitNum)
elif dataSet == 'FERET':
return loadFERET(splitNum)
elif dataSet == 'Yale':
return loadYale(splitNum)
else:
print('No this DataSet!')
y_train,
class_num,
k=(40 - 25) * split_num - 1)
# eig_vals, eig_vecs = np.linalg.eig(np.linalg.inv(S_W).dot(S_B))
eig_vals, eig_vecs = np.linalg.eig(S_B - alpha * S_W)
eig_pairs = [(np.abs(eig_vals[i]), eig_vecs[:, i])
for i in range(len(eig_vals))]
eig_pairs = sorted(eig_pairs, key=lambda k: k[0], reverse=True)
vecs = [eig_pairs[i][1].reshape(m, 1) for i in range(n_component)]
W = np.hstack(vecs)
return W
# 加载数据
split_num = 5
train_imgs, train_labels, test_imgs, test_labels = loadImg(split_num=split_num)
# 数据标准化处理
sc = StandardScaler()
X_train_std = sc.fit_transform(train_imgs)
X_test_std = sc.fit_transform(test_imgs)
y_train = np.array(train_labels)
y_test = np.array(test_labels)
# 降维
#先进行 PCA处理，以免维数过高
pca = PCA(n_components=80)
pca.fit(X_train_std)
X_train_pca = pca.transform(X_train_std)
X_test_pca = pca.transform(X_test_std)
#w = mfa(X_train_pca, y_train, X_test_pca, y_test, n_component=50)
def showORL():
train_imgs, train_labels, test_imgs, test_labels = loadImg(10)
for i in range(len(train_imgs)):
img = train_imgs[i].reshape(112,92)
plt.imshow(img,cmap='gray')
plt.show()