AttributeError: 'PreprocessClientData' object has no attribute 'create_tf_dataset_for_clients'

Question

I am doing a project in federated learning . I got this error.

  AttributeError Traceback (most recent call last)
    <ipython-input-62-c7a27f8e5bd7> in <cell line: 2>()
          1 sample_clients=emnist_train.client_ids[0:NUM_CLIENTS]
    ----> 2 federated_train_data= make_federated_data(emnist_train,sample_clients)
          3 print('Number of client datasets:{1}'.format(l=len(federated_train_data)))
         4 print('First dataset:{d}'.format(d=federated_train_data[0]))

    1 frames
    <ipython-input-60-5de6a0b9b2fa> in <listcomp>(.0)
          1 def make_federated_data(client_data,client_ids):
          2   return[
     ----> 3       preprocess(client_data.create_tf_dataset_for_clients(x))
          4       for x in client_ids
         5   ]

Can anyone help me to resolve it ?

I have tried to load different mnist dataset. But I couldn't find data with attributes create_tf_dataset_for_clients