contingency_matrix = metrics.cluster.contingency_matrix(y_true, y_pred)
# Find optimal one-to-one mapping between cluster labels and true labels
row_ind, col_ind = linear_sum_assignment(-contingency_matrix)
# Return cluster accuracy
return contingency_matrix[row_ind, col_ind].sum() / np.sum(contingency_matrix)
digits = load_digits()
data = scale(digits.data)
pca = PCA(n_components=2).fit(data)
reduced_data = PCA(n_components=2).fit_transform(data)
reduced_data1=reduced_data.copy()
plt.scatter(reduced_data[:,0],reduced_data[:,1])
plt.xlabel("X-Component")
plt.ylabel("Y-Component")
plt.show()
y_true = digits.target
print("\n\n************************Agglomerative *********************")
#----------------------------------------------------------------------------------------------#
# Data outside #
Files = pd.read_csv(FileName, header=None)
SampleNumber = Files.columns.size - 1 # Sample Number
FeatureNumber = len(Files) - 1 # Feature Number
Labels = np.array(Files.values[0, 1:SampleNumber+1], dtype='int') # Labels
Features = np.array(Files.values[1:FeatureNumber+1, 0], dtype='int') # Features
Datas = Files.iloc[1:FeatureNumber+1, 1:SampleNumber+1] # Datas
Datas = np.array(Datas.T) # sample * feature
print('Sample Number : ', SampleNumber)
print('Feature Number : ', FeatureNumber)
#----------------------------------------------------------------------------------------------#
# PCA Model #
PCAmodel = PCA()
PCAmodel.n_components = 2 # Default : None(all) , components number
PCAmodel.copy = False # Default : True, copy the origin data
PCAmodel.whiten = False # Default : False, let features with same var
PCAmodel.random_state = 0 # Default : None
PCAmodel.svd_solver = 'auto' # Default : auto
#----------------------------------------------------------------------------------------------#
# PCA Processing #
Scaler = skl.preprocessing.StandardScaler()
Scaler.mean_ = np.mean(Datas, axis=0)
Scaler.scale_ = np.std(Datas, axis=0, ddof=1)
ScaledDatas = Scaler.transform(Datas)
NewDatas = PCAmodel.fit_transform(ScaledDatas)
print('-------------------------------')
print('Pragma : PCA')
print('Reserved Comp Number : ', PCAmodel.n_components_)
print('Reserved Comp Ratio : ', PCAmodel.explained_variance_ratio_) # Larger is better
print('Reserved Comp Var : ', PCAmodel.explained_variance_) # Larger is better