utils.py

"""
Implementation of several useful tool
Please copy this script to target path
"""
import numpy as np
import random
import torch


def mnist_noniid(dataset, num_users):
    """
    Sample non-I.I.D client data from MNIST dataset
    :param dataset:
    :param num_users:
    :return:
    """
    # num_shards, num_imgs = 30, 2000
    num_shards = int(num_users*3)
    num_imgs = int(60000 / num_shards)
    idx_shard = [i for i in range(num_shards)]
    dict_users = {i: np.array([], dtype='int64') for i in range(num_users)}
    idxs = np.arange(num_shards*num_imgs)
    labels = dataset.targets.numpy()

    # sort labels
    idxs_labels = np.vstack((idxs, labels))
    idxs_labels = idxs_labels[:,idxs_labels[1,:].argsort()]
    idxs = idxs_labels[0,:]

    # divide and assign
    for i in range(num_users):
        rand_set = set(np.random.choice(idx_shard, 3, replace=False))
        idx_shard = list(set(idx_shard) - rand_set)
        for rand in rand_set:
            dict_users[i] = np.concatenate((dict_users[i], idxs[rand*num_imgs:(rand+1)*num_imgs]), axis=0)
    return dict_users


def gaussian_noise(data_shape, s, sigma, device=None):
    """
    Gaussian noise for CDP-FedAVG-LS Algorithm
    """
    return torch.normal(0, sigma * s, data_shape).to(device)