Batch generating functions

loan_data_get_data.py

from tensorflow.keras.utils import to_categorical   

#function for generating latent samples for synthetic data for generator training
def generate_latent_samples(n):
    
    #generate latent vectors with balanced targets
    x = {'latent': np.random.normal(size=(n, latent_dim)),
         'target': to_categorical(np.hstack([np.array([_x for _ in range(n//2)]) for _x in range(2)]),2)}
    
    #outputs indicating postive discirmination (target value)
    y = np.hstack([np.array([[0] for _ in range(n)]), x['target']])

    return x, y


#function for generating synthetic samples for discriminator training
def generate_synthetic_samples(n):
    
    latent_x, _ = generate_latent_samples(n)
    #generate data with balanced targets
    gen_predict = generator.predict(latent_x)

    x = dict(zip(generator.output_names, gen_predict))
    
    #outputs indicating negative discirmination
    y = np.array([[1,0,0] for _ in range(n)])

    return x, y
    
    
#function for generating real samples for discriminator training
def generate_real_samples(n):
    
    #sample real data with balanced targets
    samples = [np.random.choice(train_target_df[train_target_df[target] == _y].index, size=n//2) for _y in range(2)]

    x = {name:np.vstack([train_df[name][sample].values.reshape(-1,1) for y,sample in enumerate(samples)]) for name in data}
    for name,n_token in n_tokens.items():
        x[name] = to_categorical(x[name], n_token)
    
    #outputs indicating postive discirmination (target value)
    y = np.hstack([np.array([[0] for _ in range(n)]),
                   to_categorical(np.hstack([train_target_df[target][sample].values for sample in samples]),2)])

    return x, y