def main():
data_train, data_test, target_train, target_test = common.load_train_data_and_split(file='data/processed_missing_filled_in.csv')
data_train = np.asarray(data_train)
target_train = np.array(target_train)
target_train = target_train.astype(np.int32)
print(target_train)
data_train, target_train = smote.smote_data(data_train, target_train)
classify(data_train, target_train, data_test, target_test)
def load_train_data_and_split(testsize=0.3,
targetcol=-1,
file='data/processed_without_missing.csv',
split=True,
num_samples_per_class=-1,
num_classes=3,
smote=False):
print("Loading dataset.")
headers = []
dataset = []
ifile = open(file, "r")
reader = csv.reader(ifile)
first = True
for row in reader:
if first:
first = False
headers.append(row)
continue
dataset.append(row)
for i in row:
if not i.isdigit():
print(row)
ifile.close()
inputs = [[int(y) for y in x] for x in dataset]
outputs = []
for row in inputs:
outputs.append(row[targetcol])
del row[targetcol]
if num_samples_per_class > 0: # we want equal subsets of each class.
# first get the number of classes and their values.
output_vals = set()
for i, row in enumerate(inputs):
output_vals.add(outputs[i])
print("Number of classes: ", len(output_vals))
# then delete samples that go over the 3000 limit.
counts = [0, 0, 0]
remove_indices = []
for i, row in enumerate(inputs):
counts[
outputs[i] -
1] += 1 # mapping from target (1,2,3) to array index (0,1,2)
if (counts[outputs[i] - 1] > num_samples_per_class
): # we exceeded the count so delete this row.
remove_indices.append(i)
for i in reversed(range(len(remove_indices))):
del inputs[remove_indices[i]]
del outputs[remove_indices[i]]
print("Final counts: ", counts)
print("Num rows: ", len(inputs))
print("Done loading")
for i in range(len(outputs)):
outputs[i] -= 1
if split:
input_train, input_test, output_train, output_test = train_test_split(
inputs, outputs, test_size=testsize, random_state=42)
input_train = np.array(input_train)
input_test = np.array(input_test)
output_train = np.array(output_train)
output_train = output_train.astype(np.int32)
output_test = np.array(output_test)
output_test = output_test.astype(np.int32)
if num_classes == 2: # convert all outputs of 3 to outputs of 2.
output_train[output_train == 2] = 1
output_test[output_test == 2] = 1
if smote:
input_train, output_train = sm.smote_data(input_train,
output_train)
return input_train, input_test, output_train, output_test
else:
inputs = np.array(inputs)
outputs = np.array(outputs)
return inputs, outputs
DecisionTreeClassifier(max_depth=15),
RandomForestClassifier(max_depth=15, n_estimators=10, max_features=1),
AdaBoostClassifier(),
GaussianNB(),
LDA(),
QDA()
]
X_train, X_test, y_train, y_test = common.load_train_data_and_split(
file='data/processed_missing_filled_in.csv')
X_train = np.asarray(X_train)
y_train = np.array(y_train)
y_train = y_train.astype(np.int32)
X_train, y_train = smote.smote_data(X_train, y_train)
# iterate over classifiers
for name, clf in zip(names, classifiers):
print("Fitting " + name + "...")
predicted_test = clf.fit(X_train, y_train).predict(X_test)
test_p = ((y_test != predicted_test).sum()) / (len(X_test)) * 100
print("Error on test set: %d" % test_p)
print(metrics.classification_report(y_test, predicted_test))
'''
results (f1-score)
-> rows with missing medical speciality removed
# SVC(gamma=2, C=1),
DecisionTreeClassifier(max_depth=15),
RandomForestClassifier(max_depth=15, n_estimators=10, max_features=1),
AdaBoostClassifier(),
GaussianNB(),
LDA(),
QDA()
]
X_train, X_test, y_train, y_test = common.load_train_data_and_split(file='data/processed_missing_filled_in.csv')
X_train = np.asarray(X_train)
y_train = np.array(y_train)
y_train = y_train.astype(np.int32)
X_train, y_train = smote.smote_data(X_train, y_train)
# iterate over classifiers
for name, clf in zip(names, classifiers):
print("Fitting " + name + "...")
predicted_test = clf.fit(X_train, y_train).predict(X_test)
test_p = ((y_test != predicted_test).sum())/(len(X_test))*100
print("Error on test set: %d" % test_p)
print(metrics.classification_report(y_test, predicted_test))
'''
results (f1-score)
-> rows with missing medical speciality removed
def load_train_data_and_split(testsize=0.3, targetcol=-1, file='data/processed_without_missing.csv', split=True, num_samples_per_class=-1, num_classes=3, smote=False):
print("Loading dataset.")
headers = []
dataset = []
ifile = open(file, "r")
reader = csv.reader(ifile)
first = True
for row in reader:
if first:
first = False
headers.append(row)
continue
dataset.append(row)
for i in row:
if not i.isdigit():
print (row)
ifile.close()
inputs = [[int(y) for y in x] for x in dataset]
outputs = []
for row in inputs:
outputs.append(row[targetcol])
del row[targetcol]
if num_samples_per_class > 0: # we want equal subsets of each class.
# first get the number of classes and their values.
output_vals = set()
for i, row in enumerate(inputs):
output_vals.add(outputs[i])
print("Number of classes: ", len(output_vals))
# then delete samples that go over the 3000 limit.
counts = [0, 0, 0]
remove_indices = []
for i, row in enumerate(inputs):
counts[outputs[i]-1] += 1 # mapping from target (1,2,3) to array index (0,1,2)
if (counts[outputs[i]-1] > num_samples_per_class): # we exceeded the count so delete this row.
remove_indices.append(i)
for i in reversed(range(len(remove_indices))):
del inputs[remove_indices[i]]
del outputs[remove_indices[i]]
print("Final counts: ", counts)
print("Num rows: ", len(inputs))
print("Done loading")
for i in range(len(outputs)):
outputs[i] -= 1
if split:
input_train, input_test, output_train, output_test = train_test_split(inputs, outputs, test_size=testsize, random_state=42)
input_train = np.array(input_train)
input_test = np.array(input_test)
output_train = np.array(output_train)
output_train = output_train.astype(np.int32)
output_test = np.array(output_test)
output_test = output_test.astype(np.int32)
if num_classes == 2: # convert all outputs of 3 to outputs of 2.
output_train[output_train == 2] = 1
output_test[output_test == 2] = 1
if smote:
input_train, output_train = sm.smote_data(input_train, output_train)
return input_train, input_test, output_train, output_test
else:
inputs = np.array(inputs)
outputs = np.array(outputs)
return inputs, outputs
"ID3:",
id3(X_train_r_abalone, Y_train_r_abalone, X_test_r_abalone,
Y_test_r_abalone))
print(
"ANN:",
neural_net(X_train_r_abalone,
Y_train_r_abalone,
X_test_r_abalone,
Y_test_r_abalone,
0.0001,
sigmoidal,
10000,
prt=False))
print("\nAbalone smoted:")
x_abalone_new, y_abalone_new = sm.smote_data(x_abalone_r, y_abalone_r, dts, 5)
X_train_abalone_new, Y_train_abalone_new, X_test_abalone_new, Y_test_abalone_new = separate(
0.3, x_abalone_new, y_abalone_new, S)
print(
"NBG:",
nb_gaussiano(X_train_abalone_new, Y_train_abalone_new, X_test_abalone_new,
Y_test_abalone_new))
print(
"KNN:",
knn(X_train_abalone_new, Y_train_abalone_new, X_test_abalone_new,
Y_test_abalone_new, 15))
print(
"ID3:",
id3(X_train_abalone_new, Y_train_abalone_new, X_test_abalone_new,
Y_test_abalone_new))
print(
