def prob0():
arff = Arff('datasets/labor.arff', label_count=1)
# Trim the id column
arff = arff.create_subset_arff(col_idx=slice(1, None))
arff = arff.get_features()
km = KMeans(5)
km.train(arff, verbose=True, centers=arff.data[:5])
def test_load_arff(self):
""" Tests downloading and loading arff file
"""
t = Arff()
t.load_arff(self.iris_path)
self.assertListEqual(t.data[t.shape[0]-1].tolist(), [5.9, 3.0, 5.1, 1.8, 2.0])
def prob_0():
arff = Arff('datasets/lenses.arff')
d = DecisionTreeLearner()
f = arff.get_features()
l = arff.get_labels()
d.train(f,l)
print(d.tree)
def prob0haccomplete():
arff = Arff('datasets/labor.arff', label_count=1)
# Trim the id column
arff = arff.create_subset_arff(col_idx=slice(1, None))
arff = arff.get_features()
hac = HAC(simple=False)
hac.train(arff, verbose=True, printk=[5])
def setUp(self):
path = os.path.join(utils.get_root(), "test/datasets/cm1_req.arff")
data = Arff(arff=path)
self.features = data.get_features()
self.labels = data.get_labels()
self.learner = BaselineLearner()
def setUp(self):
# NOTE: for discrete attributes, at least one value must be a float in order for numpy array
# functions to work properly.
data = np.array([[1.5, -6, 1.0],
[2.3, -8, 2],
[4.1, self.infinity, 2]])
m = Arff(data, label_count=1)
m.attr_names = ['A', 'B', 'C']
m.str_to_enum = [{}, {}, {'R': 0, 'G': 1, 'B': 2}]
m.enum_to_str = [{}, {}, {0: 'R', 1: 'G', 2: 'B'}]
self.m = m
data2 = np.array([[0.0, 1.0, 2.0, 3.0, 0.0],
[0.1, 1.1, 2.1, 3.1, 1.0],
[0.2, 1.2, 2.2, 3.2, 1.0],
[0.3, 1.3, 2.3, 3.3, 2.0],
[0.4, 1.4, 2.4, 3.4, 2.0]])
m2 = Arff(data2, label_count=1)
m2.attr_names = ['A', 'B', 'C', 'D', 'E']
m2.str_to_enum = [{}, {}, {}, {}, {'R': 0, 'G': 1, 'B': 2}]
m2.enum_to_str = [{}, {}, {}, {}, {0: 'R', 1: 'G', 2: 'B'}]
self.m2 = m2
self.credit_data_path = os.path.join(utils.get_root(),"test/datasets/creditapproval.arff")
self.iris_path = os.path.join(utils.get_root(),"test/datasets/iris.arff")
def main():
arff = Arff(sys.argv[1])
pl = PerceptronLearner()
features = arff.get_features()
labels = arff.get_labels()
accuracy_matrix = np.zeros((5, 20))
for i in range(5):
pl.train(features, labels)
a = pl.accuracy_tracker[:20]
# pad to make 20 wide
a = np.pad(a, (0, 20 - len(a)), 'constant', constant_values=a[-1])
accuracy_matrix[i] = a
# Average the accuracies of each step
print(accuracy_matrix)
avg_accuracy = np.sum(accuracy_matrix, axis=0) / 5
print(avg_accuracy)
plt.plot(1 - avg_accuracy)
plt.xlabel("Epochs")
plt.ylabel("Avg Misclassification Rate")
plt.title("Avg Misclassification Rate Over Epochs")
plt.show()
def main():
arff = Arff(sys.argv[1])
features = arff.get_features()
labels = arff.get_labels()
pl = PerceptronLearner()
pl.train(features, labels)
visualize_training(features, labels, pl)
def setup():
arff = Arff('datasets/labor.arff', label_count=1)
# Trim the id column
arff = arff.create_subset_arff(col_idx=slice(1, None))
arff = arff.get_features()
hac = HAC()
hac.nominal_indicies = np.where(np.array(arff.attr_types) == 'nominal')[0]
print('33,44', hac.get_distance(arff.data[33], arff.data[44]))
print('25,34', hac.get_distance(arff.data[25], arff.data[34]))
def test_arff_constructor(self):
""" Tests construction of Arff from path, arff, numpy array
"""
## Create a Matrix object from arff
credit = Arff(arff=self.credit_data_path)
credit3 = Arff(arff=credit)
credit2 = Arff(arff=credit.data)
np.testing.assert_array_almost_equal(credit.data, credit2.data)
np.testing.assert_array_almost_equal(credit2.data, credit3.data)
def prob4h():
arff = Arff('datasets/abalone.arff', label_count=0)
arff.normalize()
domain = np.arange(2, 8)
print('single link --------------------')
hoc = HAC()
hoc.train(arff, printk=domain, silhouette=True)
print('complete link -----------------------')
hoc = HAC(simple=False)
hoc.train(arff, printk=domain, silhouette=True)
def prob4():
arff = Arff('datasets/abalone.arff', label_count=0)
arff.normalize()
domain = np.arange(2, 8)
ssekmm = []
for k in domain:
km = KMeans(k)
ssek = km.train(arff)
ssekmm.append(ssek)
print(km.calc_silhouette_score())
def prob2():
iris = Arff('datasets/iris.arff')
features = iris.get_features()
# features.normalize()
# Train k means for 2-7
ks = [2, 3, 4, 5, 6, 7]
for k in ks:
km = KMeans(k)
km.train(features)
hac2 = HAC(simple=False)
hac2.train(features, printk=ks)
def test_create_subset_arff(self):
m2 = Arff(self.m2, [1,2], slice(1,3))
self.assertEqual(m2.shape, (2,2))
m2 = Arff(self.m2, slice(1,3), [1,2])
self.assertEqual(m2.shape, (2,2))
# Automatic label inference
self.m2.label_count=3
m2 = Arff(self.m2, slice(1,3), slice(1,None), label_count=None)
self.assertEqual(3, m2.label_count)
m2 = Arff(self.m2, slice(1,3), slice(1,-1), label_count=None)
self.assertEqual(2, m2.label_count)
def prob5():
arff = Arff(sys.argv[2])
imp_atts = [1, 3, 4, 5, 7, 9, 11, 12, 13]
arff.shuffle()
n = len(arff.get_labels().data)
t = int(n * .55)
v = n - int(n * .20)
train_set = arff.create_subset_arff(row_idx=slice(0, t, 1),
col_idx=imp_atts)
test_set = arff.create_subset_arff(row_idx=slice(t, v, 1),
col_idx=imp_atts)
validation_set = arff.create_subset_arff(row_idx=slice(v, n, 1),
col_idx=imp_atts)
epochs = []
momentums = np.linspace(0, 1.5, 20)
# momentums = [.5, 1]
for momentum in momentums:
print(momentum)
nn = NeuralNetwork(8, [30], 11, LR=.1, momentum=momentum)
all_acc_va, all_mse_va, all_mse_te, all_mse_tr = nn.train_set(
train_set, test_set, validation_set, w=5)
epochs.append(len(all_acc_va))
plt.plot(momentums, epochs)
plt.title("Vowel Momentum vs Epoch Convergence")
plt.xlabel("Momentum")
plt.ylabel("Epochs til Conv.")
plt.show()
def test_copy_and_slice(self):
d = Arff(self.credit_data_path)
e = d.copy()
e._copy_and_slice_arff(d, 1, 5)
self.assertEqual(e.shape, (1,1))
e._copy_and_slice_arff(d, slice(1,4), slice(2,4))
self.assertEqual(e.shape, (3, 2))
# This will create a 1D array, returning coords (1,1), (2,5), (3,7)
e._copy_and_slice_arff(d, [1,2,3,7], [1,5,7,8])
self.assertEqual(e.shape, (4, ))
e._copy_and_slice_arff(d, [1,2,3,7], slice(0,5))
self.assertEqual(e.shape, (4, 5))
def test_append_columns(self):
test_matrix = Arff(self.m)
# Verify it works with other matrices
test_matrix.append_columns(test_matrix)
assert test_matrix.data.shape == (3,6)
# Verify it works with numpy array
test_matrix.append_columns(test_matrix.data)
assert test_matrix.data.shape == (3,12)
# Verify it works with 2D list
test_matrix.append_columns(test_matrix.data)
assert test_matrix.data.shape == (3,24)
# Verify incompatible number of columns
with self.assertRaises(Exception) as context:
test_matrix.append_columns(self.m.data[:1,:])
self.assertTrue('Incompatible number of rows' in str(context.exception))
def test_append_rows(self):
test_matrix = Arff(self.m)
# Verify it works with other matrices
test_matrix.append_rows(test_matrix)
assert test_matrix.data.shape == (6,3)
# Verify it works with numpy array
test_matrix.append_rows(test_matrix.data)
assert test_matrix.data.shape == (12,3)
# Verify it works with 2D list
test_matrix.append_rows(test_matrix.data)
assert test_matrix.data.shape == (24,3)
# Verify incompatible number of rows
with self.assertRaises(Exception) as context:
test_matrix.append_rows(self.m.data[:,:2])
print(str(context.exception))
self.assertTrue('Incompatible number of columns' in str(context.exception))
def prob2wclass():
iris = Arff('datasets/iris.arff', label_count=0)
# features.normalize()
# Train k means for 2-7
ks = [2, 3, 4, 5, 6, 7]
for k in ks:
km = KMeans(k)
km.train(iris)
hac2 = HAC(simple=False)
hac2.train(iris, printk=ks)
def prob_4(lr):
# read in vowels dataset
arff = Arff('datasets/vowels.arff')
# Leave out the test/train and person features, which are unceccessary.
arff = arff.create_subset_arff(col_idx=slice(2, None), label_count=1)
# Get a 75/25 split
arff.shuffle()
training = arff.create_subset_arff(slice(arff.instance_count // 4))
test = arff.create_subset_arff(slice(arff.instance_count // 4, -1))
t_features = test.get_features()
t_labels = test.get_labels()
# Get a 15% Validation set
validation = training.create_subset_arff(slice(arff.instance_count // 5))
training = training.create_subset_arff(
slice(arff.instance_count // 5, None))
v_features = validation.get_features()
v_labels = validation.get_labels()
domain = 2**np.arange(0, 8)
training_mse = []
validation_mse = []
test_mse = []
for nodes in domain:
mse = 0
vmse = 0
tmse = 0
for _ in range(100):
learner = MultilayerPerceptronLearner(
[training.features_count, nodes, 11], momentum=0)
# learner.zero_weights()
learner.lr = lr
learner.max_epoch = 500
learner.train(training.get_features(), training.get_labels(),
validation.get_features(), validation.get_labels())
tmse += learner.get_mse(test.get_features(), test.get_labels())
mse += learner.get_mse(training.get_features(),
training.get_labels())
vmse += learner.get_mse(validation.get_features(),
validation.get_labels())
training_mse.append(mse / 100)
validation_mse.append(vmse / 100)
test_mse.append(tmse / 100)
plt.semilogx(domain, test_mse, basex=2, label="Test Set MSE")
plt.semilogx(domain, training_mse, basex=2, label="Training Set MSE")
plt.semilogx(domain, validation_mse, basex=2, label="Validation Set MSE")
plt.title("MSE vs Number of Hidden Nodes")
plt.xlabel("Number of Hidden Nodes")
plt.ylabel("Mean Squared Error")
plt.legend()
plt.show()
def prob_6():
# Repeat experiments for magic telescope and housing using weights (w = 1/dist**2)
arff = Arff('datasets/credit.arff')
arff.shuffle()
test = arff.create_subset_arff(slice(arff.instance_count // 4))
train = arff.create_subset_arff(slice(arff.instance_count // 4, None))
train.normalize()
test.normalize()
krange = np.arange(1, 16, 2)
accs = []
for k in krange:
knn = KNN(k, weighting=True, vdm=True)
predictions = knn.knn(train.get_features(), train.get_labels(),
test.get_features())
acc = predictions == np.ravel(test.get_labels().data)
print("k:", k, "accuracy:", sum(acc) / len(acc))
accs.append(sum(acc) / len(acc))
plt.plot(krange, accs)
plt.title("K Size Versus Accuracy on Credit Approval (Weighted)")
plt.xlabel("K")
plt.ylabel("Accuracy")
plt.show()
def prob_1():
arff = Arff('datasets/lenses.arff')
acc = []
for _ in range(10):
arff.shuffle()
testing = arff.create_subset_arff(slice(arff.instance_count//5))
training = arff.create_subset_arff(slice(arff.instance_count//5, None))
d = DecisionTreeLearner()
features = training.get_features()
labels = training.get_labels()
t_feat = testing.get_features()
t_labels = testing.get_labels()
d.train(features, labels)
accuracy = d.get_accuracy(t_feat, t_labels)
print(accuracy)
acc.append(accuracy)
print(sum(acc)/len(acc))
def test_cases():
# test_1()
attr_types = [
"real", "real", "real", "real", "cat", "real", "cat", "real", "real",
"cat", "real", "cat", "cat", "cat", "cat", "cat", "cat"
]
attr_idx = [[], [], [], [], ['none', 'tcf', 'tc'], [],
['none', 'ret_allw', 'empl_contr'], [], [], ['yes', 'no'], [],
['below_average', 'average', 'generous'], ['yes', 'no'],
['none', 'half', 'full'], ['yes', 'no'],
['none', 'half', 'full'], ['bad', 'good']]
k = 5
arff = Arff("labor.arff")
# arff.normalize()
features = arff.get_features().data
labels = arff.get_labels().data
# attributes = arff.get_attr_names()
data = np.hstack((features, labels))[:, 1:-1]
kmc = HA_Clustering(k, data, data, attr_types, "complete link", attr_idx)
def prob_5():
cont_mask = [1, 2, 7, 10, 13, 14, 16]
cate_mask = [0, 3, 4, 5, 6, 8, 9, 11, 12, 15]
arff = Arff("credit_approval_data.arff")
arff.shuffle()
arff.normalize()
n = len(arff.get_labels().data)
t = int(n * .7)
train_data = arff.create_subset_arff(row_idx=slice(0, t, 1))
test_data = arff.create_subset_arff(row_idx=slice(t, n, 1))
test_data = np.hstack((test_data.get_features().data, test_data.get_labels().data))
train_data = np.hstack((train_data.get_features().data, train_data.get_labels().data))
#b,30.83,0,u,g,w,v,1.25,t,t,01,f,g,00202,0,+
dist_matrix = np.ones((16, 16))
np.fill_diagonal(dist_matrix, 0)
KNNC = KNNClassifier(8, train_data, test_data)
print(KNNC.get_accuracy_mixed(cate_mask, cont_mask, dist_matrix))
def main():
# Train each perceptron on its own split data set
fast_v_mid = Arff('datasets/restaurants/fast_v_mid.arff')
fast_v_fine = Arff('datasets/restaurants/fast_v_fine.arff')
mid_v_fine = Arff('datasets/restaurants/mid_v_fine.arff')
pl_fast_mid = PerceptronLearner()
pl_fast_fine = PerceptronLearner()
pl_mid_v_fine = PerceptronLearner()
# Train each perceptron
train_perceptron(pl_fast_mid, fast_v_mid)
train_perceptron(pl_fast_fine, fast_v_fine)
train_perceptron(pl_mid_v_fine, mid_v_fine)
# Run on new data
# Burger King
burger_king = [4, 2, 2]
# Cheesecake Factory
cheesecake_factory = [2, 4, 4]
# Best fine-dining in the world
best_fine_dining = [1, 5, 3]
print_findings(
"Burger King",
determine_category(burger_king, pl_fast_mid, pl_fast_fine,
pl_mid_v_fine))
print_findings(
"Cheesecake Factory",
determine_category(cheesecake_factory, pl_fast_mid, pl_fast_fine,
pl_mid_v_fine))
print_findings(
"'Best Food In The World'",
determine_category(best_fine_dining, pl_fast_mid, pl_fast_fine,
pl_mid_v_fine))
def prob3():
""" """
arff = Arff(sys.argv[2])
imp_atts = [1, 3, 4, 5, 7, 9, 11, 12, 13]
arff.shuffle()
n = len(arff.get_labels().data)
t = int(n * .55)
v = n - int(n * .20)
train_set = arff.create_subset_arff(row_idx=slice(0, t, 1),
col_idx=imp_atts)
test_set = arff.create_subset_arff(row_idx=slice(t, v, 1),
col_idx=imp_atts)
validation_set = arff.create_subset_arff(row_idx=slice(v, n, 1),
col_idx=imp_atts)
best_mse_te = []
best_mse_tr = []
best_mse_va = []
epochs = []
LRS = [.01, .1, .5, .8, 1.5]
for LR in LRS:
# print(LR)
nn = NeuralNetwork(8, [16], 11, LR=LR, momentum=0)
all_acc_va, all_mse_va, all_mse_te, all_mse_tr = nn.train_set(
train_set, test_set, validation_set, w=5)
best_mse_te.append(min(all_mse_te))
best_mse_tr.append(min(all_mse_tr))
best_mse_va.append(min(all_mse_va))
epochs.append(len(all_mse_va))
plt.plot(LRS, best_mse_te, label="MSE Te")
plt.plot(LRS, best_mse_tr, label="MSE Tr")
plt.plot(LRS, best_mse_va, label="MSE V.A")
plt.title("Vowel MSE vs Learning Rate")
plt.xlabel("Learning Rate")
plt.ylabel("MSE")
plt.legend()
plt.show()
plt.plot(LRS, epochs)
plt.title("Vowel Epochs vs Learning Rate")
plt.xlabel("Learning Rate")
plt.ylabel("Epochs")
plt.legend()
plt.show()
def prob3_normalized():
arff = Arff('datasets/abalone.arff', label_count=0)
arff.normalize()
domain = np.arange(2, 8)
ssekmm = []
for k in domain:
km = KMeans(k)
ssek = km.train(arff)
ssekmm.append(ssek)
hac = HAC()
hac2 = HAC(simple=False)
ssehac = hac.train(arff, printk=domain)
ssehac2 = hac2.train(arff, printk=domain)
plt.plot(domain, ssekmm, label="K-Means SSE")
plt.plot(domain, ssehac[::-1], label="HAC (Single-Link) SSE")
plt.plot(domain, ssehac2[::-1], label="HAC (Complete-Link) SSE")
plt.title("Abalone SSE (Normalized) vs # of Clusters")
plt.xlabel("# of Clusters")
plt.ylabel('SSE')
plt.legend()
plt.show()
def prob_6_b():
# read in vowels dataset
arff = Arff('datasets/vowels.arff')
# Leave out the test/train and person features, which are unceccessary.
arff = arff.create_subset_arff(col_idx=slice(2, None), label_count=1)
# Get a 75/25 split
arff.shuffle()
training = arff.create_subset_arff(slice(arff.instance_count // 4))
test = arff.create_subset_arff(slice(arff.instance_count // 4, -1))
t_features = test.get_features()
t_labels = test.get_labels()
# Get a 15% Validation set
validation = training.create_subset_arff(slice(arff.instance_count // 5))
training = training.create_subset_arff(
slice(arff.instance_count // 5, None))
v_features = validation.get_features()
v_labels = validation.get_labels()
features = training.get_features()
labels = training.get_labels()
taccuracy = []
accuracy = []
domain = 2**np.arange(1, 7)
for i in domain:
tacc = 0
acc = 0
for _ in range(3):
learner = MultilayerPerceptronLearner([training.features_count] +
[i] * (32 // i) + [11],
momentum=.85)
# learner.zero_weights()
learner.lr = .1
learner.max_epoch = 500
learner.train(training.get_features(), training.get_labels(),
validation.get_features(), validation.get_labels())
tacc += learner.get_accuracy(t_features, t_labels)
acc += learner.get_accuracy(features, labels)
accuracy.append(acc / 3)
taccuracy.append(tacc / 3)
plt.semilogx(domain, accuracy, basex=2, label="Training Set Accuracy")
plt.semilogx(domain, taccuracy, basex=2, label="Test Set Accuracy")
plt.title("Node Distribution vs Accuracy")
plt.xlabel("Number of Nodes per Hidden Layer")
plt.ylabel("Accuracy")
plt.legend()
plt.show()
def prob_5(lr, hidden):
# read in vowels dataset
arff = Arff('datasets/vowels.arff')
# Leave out the test/train and person features, which are unceccessary.
arff = arff.create_subset_arff(col_idx=slice(2, None), label_count=1)
# Get a 75/25 split
arff.shuffle()
training = arff.create_subset_arff(slice(arff.instance_count // 4))
test = arff.create_subset_arff(slice(arff.instance_count // 4, -1))
t_features = test.get_features()
t_labels = test.get_labels()
# Get a 15% Validation set
validation = training.create_subset_arff(slice(arff.instance_count // 5))
training = training.create_subset_arff(
slice(arff.instance_count // 5, None))
v_features = validation.get_features()
v_labels = validation.get_labels()
epochs = []
accuracy = []
domain = np.linspace(0, 1, 20)
for momentum in domain:
e = 0
acc = 0
for _ in range(10):
learner = MultilayerPerceptronLearner(
[training.features_count, hidden, 11], momentum=momentum)
# learner.zero_weights()
learner.lr = lr
learner.max_epoch = 500
learner.train(training.get_features(), training.get_labels(),
validation.get_features(), validation.get_labels())
acc += learner.get_accuracy(t_features, t_labels)
e += learner.epochs
epochs.append(e / 10)
accuracy.append(acc / 10)
print(accuracy)
plt.plot(domain, epochs)
plt.title("Number of Training Epochs vs Momentum")
plt.xlabel("Momentum Constant")
plt.ylabel("Number of Training Epochs")
plt.show()
def prob_2():
# Get accuracy on cars.arff
arff = Arff('datasets/cars.arff')
arff.shuffle()
acc, tacc, = k_fold_cv(arff, 10)
print('cars:')
print('acc',acc)
print('tacc',tacc)
print('tot',sum(tacc)/len(tacc))
print()
# Get accuracy of voting.arff
arff = Arff('datasets/voting.arff')
arff.shuffle()
acc,tacc, = k_fold_cv(arff, 10)
print('voting;')
print('acc',acc)
print('tacc',tacc)
print('tot',sum(tacc)/len(tacc))
print()