class_counter2.most_common(1)[0][0], "Actual", afaf)
return class_counter2.most_common(1)[0][0]
if __name__ == '__main__':
np.random.seed(1)
df = pd.read_csv(data_filename)
print(df.shape)
# data_training = df.iloc[:, 1:17]
# gsom = GSOM(.83, 16, max_radius=4)
# gsom.fit(data_training.to_numpy(), 100, 50)
# x= (data_training.to_numpy())
# gsom.predict(df,"Name","label")
X, y = pp.preProcess(data_filename)
X_f, y_f = GSMOTE.OverSample(X, y)
y_f = y_f.astype(int)
y1 = np.copy(y_f)
y = np.column_stack([y1, y_f])
labels = ["Name", "label"]
y = np.vstack((labels, y))
frame = pd.DataFrame(y[1:, :], columns=y[0, :])
gsom1 = GSOM(.83, X_f.shape[1], max_radius=4)
gsom1.fit(X_f[:-10, :], 100, 50)
gsom1.labelling_gsom(X_f[:-10, :], frame.iloc[:-10, :], "Name", "label")
gsom1.finalize_gsom_label()
y_pred = gsom1.predict_values(X_f[-10:, :], frame.iloc[-10:, :])
print(y_pred)
def predict(self, X):
return self.gsom.predict_values(X)
# def score(self, X, y=None):
# # counts number of values bigger than mean
# return(sum(self.predict(X)))
#
from sklearn.model_selection import GridSearchCV, train_test_split
date_file = "../../data/adultmini.csv".replace('\\', '/')
# date_file = "content/pygsom/data/adult.csv".replace('\\','/')
X, y = pp.preProcess(date_file)
# try different combination of hyper paramenters
parameters = [{
'smooth_iteration': [12, 25],
'training_iteration': [25, 50],
'spreading_factor': [0.83, 0.53, 0.7],
'FD': [0.1, 0.05, 0.2],
'learning_rate': [0.3, 0.4, 0.5],
'smooth_learning_factor': [0.8, 0.6, 0.7]
}]
gs = GridSearchCV(MeanClassifier(), parameters)
gs.fit(X, y)
params = gs.best_params_
print(params)
def parse_input_zoo_data(filename, header='infer'):
gsmote = GeometricSMOTE(random_state=1)
#
# (X_train, y_train), (X_test, y_test) = mnist.load_data()
#
# d1, d2, d3 = X_train.shape
# X_train_reshaped = X_train.reshape(d1, d2 * d3)
# print(X_train_reshaped[:2000, :].shape)
# y_train_half = y_train[:2000]
# classes = y_train_half.tolist()
# labels = y_train_half.tolist()
# # print(labels)
#
# input_database = {
# 0: X_train_reshaped[:2000, :]
# }
#GSMOTE
# X_f,y_f = GSMOTE.OverSample()
#
#
# X_t, X_test, y_t, y_test = train_test_split(X_f, y_f, test_size=0.2, random_state=0)
#
#
# classes = y_t.tolist()
# labels = y_t.tolist()
# input_database = {
# 0: X_t
# }
X, y = pp.preProcess(filename)
X_t, X_test, y_t, y_test = train_test_split(X,
y,
test_size=0.2,
random_state=0)
X_train, y_train = gsmote.fit_resample(X_t, y_t)
classes = y_train.tolist()
labels = y_train.tolist()
input_database = {0: X_train}
# (X_train, y_train), (X_test, y_test) = mnist.load_data()
#
# d1, d2, d3 = X_train.shape
# X_train_reshaped = X_train.reshape(d1, d2 * d3)
# print(X_train_reshaped[:2000, :].shape)
# y_train_half = y_train[:2000]
# classes = y_train_half.tolist()
# labels = y_train_half.tolist()
# # print(labels)
#
# input_database = {
# 0: X_train_reshaped[:2000, :]
# }
#Smote
# X_f,y_f = smote.Data_Extract(filename)
# classes = y_f.tolist()
# labels = y_f.tolist()
# input_database = {
# 0: X_f[:,:]
# }
# input_data = pd.read_csv(filename, header=header)
#
# input_database = {
# 0: input_data.as_matrix([0,1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,17,18,19,20,21,22,23,24,25,26,27,28,29])
# }
#
# (X_train, y_train), (X_test, y_test) = mnist.load_data()
#
# d1, d2, d3 = X_train.shape
# X_train_reshaped = X_train.reshape(d1, d2 * d3)
# print(X_train_reshaped[:2000, :].shape)
# y_train_half = y_train[:2000]
# classes = y_train_half.tolist()
# labels = y_train_half.tolist()
# # print(labels)
#
# input_database = {
# 0: X_train_reshaped[:2000, :]
# }
# input_data = pd.read_csv(filename, header=header)
#
# classes = input_data[17].tolist()
# labels = input_data[0].tolist()
# input_database = {
# 0: input_data.as_matrix([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16])
# }
return input_database, labels, classes, X_test, y_test
