for mod in range(model_no):
sheets[ind].write(mod+1,0,'model_{}'.format(mod))
for train_index,test_index in skf.split(atribute, output):
lst = [
models.Fair_rew_RF(un_gr, pr_gr),
models.FAD_class(input_size = inp, num_layers_z = 2, num_layers_y = 2,
step_z = 1.5, step_y = 1.5),
models.FAIR_scalar_class(input_size = inp, num_layers_w = 2, step_w = 1.5,
num_layers_A = 1, step_A = 1.5, num_layers_y = 2, step_y = 1.5),
models.FAIR_betaSF_class(input_size = inp, num_layers_w = 2, step_w = 1.5,
num_layers_A = 1, step_A = 1.5, num_layers_y = 2, step_y = 1.5),
models.FAIR_Bernoulli_class(input_size = inp, num_layers_w = 2, step_w = 1.5,
num_layers_A = 1, step_A = 1.5, num_layers_y = 2, step_y = 1.5),
models.FAIR_betaREP_class(input_size = inp, num_layers_w = 2, step_w = 1.5,
num_layers_A = 1, step_A = 1.5, num_layers_y = 2, step_y = 1.5),
models.FAD_prob_class(flow_length = 2, no_sample = 32,
input_size = inp, num_layers_y = 2,
step_y = 2, step_z = 2)]
x_train, x_test = atribute.iloc[train_index,:], atribute.iloc[test_index,:]
y_train, y_test = output.iloc[train_index], output.iloc[test_index]
A_train, A_test = sensitive.iloc[train_index], sensitive.iloc[test_index]
data_train, data_test = data.iloc[train_index,:], data.iloc[test_index,:]
if std_scl == 1:
std_scl = StandardScaler()
std_scl.fit(x_train)
x_train_x = std_scl.transform(x_train)
step_w=2,
num_layers_A=2,
step_A=2,
num_layers_y=3,
step_y=2),
models.FAIR_Bernoulli_class(input_size=inp,
num_layers_w=2,
step_w=2,
num_layers_A=2,
step_A=2,
num_layers_y=3,
step_y=2),
models.FAIR_betaREP_class(input_size=inp,
num_layers_w=2,
step_w=2,
num_layers_A=2,
step_A=2,
num_layers_y=3,
step_y=2),
models.FAD_prob_class(flow_length=2,
no_sample=1,
input_size=inp,
num_layers_y=3,
step_y=2,
step_z=2)
]
x_train, x_test, y_train, y_test, A_train, A_test, data_train, data_test = train_test_split(
atribute, output, sensitive, data, test_size=0.2, random_state=42)
if std_scl == 1:
