Image loading from local directory

Question

tydef prepare_data(batch_size):
(X_train, y_train)=load_data(TRAIN_DIR)
(X_test, y_test) = load_data(TEST_DIR)
X_all = np.concatenate([X_train, X_test])
y_all = np.concatenate([y_train, y_test])
X_all = X_all.astype(np.float32) / 255
X_all = X_all.reshape(-1, 28, 28, 1) * 2. - 1.
y_all = keras.utils.to_categorical(y_all, 10)

dataset = tf.data.Dataset.from_tensor_slices((X_all, y_all))
dataset = dataset.shuffle(1024)
dataset = dataset.batch(BATCH_SIZE, drop_remainder=True).prefetch(1)pe here
return dataset

This is the script to load the directory files using TRAIN_DIR variable.
But when I call the function dataset = prepare_data(BATCH_SIZE) it says "too many values to unpack (expected 2)".
Can you share your experiences?

Which line is giving you this error ? It could come from the first and second line where you want 2 output from `load data` and the function gives you only 1. — Till, Feb 17 '23 at 14:14
1 def prepare_data(batch_size): ----> 2 (X_train, y_train)=load_data(TRAIN_DIR) 3 (X_test, y_test) = load_data(TEST_DIR) 4 X_all = np.concatenate([X_train, X_test]) 5 y_all = np.concatenate([y_train, y_test]) — Mulugeta Shitie, Feb 17 '23 at 14:16
And where does your method `load_data` comes from ? What should it be returning ? — Till, Feb 17 '23 at 14:19
dataset = prepare_data(BATCH_SIZE) gan = ConditionalGAN( discriminator=discriminator, generator=generator, latent_dim=CODINGS_SIZE ) gan.compile( d_optimizer=keras.optimizers.Adam(learning_rate=0.0001), g_optimizer=keras.optimizers.Adam(learning_rate=0.0001), loss_fn=keras.losses.BinaryCrossentropy(from_logits=True) ) — Mulugeta Shitie, Feb 17 '23 at 14:27
Could you copy paste the `load_data` function that you use in the first line please? — Till, Feb 17 '23 at 14:40
def load_data(dir_path, img_size=(100,100)): """ Load resized images as np.arrays to workspace """ X = [] y = [] i = 0 label = dict() — Mulugeta Shitie, Feb 17 '23 at 14:50
TRAIN_DIR = '/content/drive/MyDrive/computer_vision/main_dataset5/train/' def prepare_data(batch_size): (X_train, y_train,X_test, y_test)=load_data(TRAIN_DIR) — Mulugeta Shitie, Feb 17 '23 at 14:51
dataset = prepare_data(BATCH_SIZE) gan = ConditionalGAN( discriminator=discriminator, generator=generator, latent_dim=CODINGS_SIZE ) — Mulugeta Shitie, Feb 17 '23 at 14:51
I don't see a `return` statement in your `load_data`. It could be the problem... — Till, Feb 17 '23 at 14:55
X = np.array(X) y = np.array(y) print(f'{len(X)} images loaded from {dir_path} directory.') return X, y, label — Mulugeta Shitie, Feb 17 '23 at 15:12
def prepare_data(batch_size): X_train, y_train, lable = load_data(TRAIN_DIR) y_all = np.concatenate([y_train, y_test]) X_all = X_all.astype(np.float32) / 255 — Mulugeta Shitie, Feb 17 '23 at 17:18
If the error is different then this issue is solved. You should validate it and then ask another question. My advice would be to give all the element in the question (all the functions) and also the complete error trace to avoid all these comments. — Till, Feb 17 '23 at 18:03
Here is maybe an answer to the other error: https://stackoverflow.com/questions/4674473/valueerror-setting-an-array-element-with-a-sequence — Till, Feb 17 '23 at 18:06
mnist = datasets.MNIST(root='../data/', train=True, transform=transform, download=True) data_loader = DataLoader(dataset=mnist, batch_size=batch_size, shuffle=True, drop_last=True) — Mulugeta Shitie, Feb 18 '23 at 13:32
how to change this file loader to my path to load from my device — Mulugeta Shitie, Feb 18 '23 at 13:33
Please, validate this answer and ask another question. This help other to find solution to the same problem as yours. — Till, Feb 28 '23 at 11:18

score 0 · Answer 1 · answered Feb 17 '23 at 15:17

Based on the comments, you have a function load_data like this:

def load_data(dir_path, img_size=(100,100)):
    """ Load resized images as np.arrays to workspace """
    X = []
    y = []
    i = 0
    label = dict()
    X = np.array(X)
    y = np.array(y)
    print(f'{len(X)} images loaded from {dir_path} directory.')
    return X, y, label

which return two numpy arrays and one dictionary.

So I would change the beginning of the function prepare_data like so:

def prepare_data(batch_size):
   X_train, y_train, label_train = load_data(TRAIN_DIR)
   X_test, y_test, label_test = load_data(TEST_DIR)

to match load_data signature.

Image loading from local directory

1 Answers1