qfnn.qf_fb.c_input module

class qfnn.qf_fb.c_input.ToQuantumData(img_size)

Bases: object

class ToQuantumData is to transform the input image into a unitary matrix to be encoded in quantum unitary matrix.

class qfnn.qf_fb.c_input.ToQuantumData_Batch

Bases: object

qfnn.qf_fb.c_input.load_data(interest_num, datapath, isppd, img_size, batch_size, inference_batch_size, is_to_q=True, num_workers=0)

Function load_data is to get the DataLoader of MNIST after preprocessing with the interest nums.

Parameters

  • interest_num – the specfic label list,such as [3,6]

  • datapath – the path of MNIST data

  • isppd – whether the data is loaded from qfMNIST,whch have done preprocessing

  • img_size – the width /height of input image (width = height)

  • batch_size – batch size while training

  • inference_batch_size – batch size while inferencing

  • is_to_q – whether the output data in loader is to be encoded in quantum unitary matrix

  • num_workers – for torch.utils.data.DataLoader.

Returns

a torch.utils.data.DataLoader for training. test_loader: a torch.utils.data.DataLoader for testing.

Return type

train_loader

qfnn.qf_fb.c_input.modify_target(target, interest_num)
qfnn.qf_fb.c_input.modify_target_ori(target, interest_num)
qfnn.qf_fb.c_input.select_num(dataset, interest_num)

Function select_num is to select the specfic label of a dataset to generate a sub-dataset. :param dataset: a pytorch-datasets :param interest_num: the specfic label list,such as [3,6]

Returns

a sub-dataset

Return type

dataset

qfnn.qf_fb.c_input.to_quantum_matrix(tensor)

Function to_quantum_matrix is to transform the input image into a unitary matrix. :param tensor: input image

Returns

a unitary matrix.

Return type

output_matrix