ncullen93

def transform_matrix_offset_center(matrix, x, y):
	o_x = float(x) / 2 + 0.5
	o_y = float(y) / 2 + 0.5
	offset_matrix = np.array([[1, 0, o_x], [0, 1, o_y], [0, 0, 1]])
	reset_matrix = np.array([[1, 0, -o_x], [0, 1, -o_y], [0, 0, 1]])
	transform_matrix = np.dot(np.dot(offset_matrix, matrix), reset_matrix)
	return transform_matrix
	
def apply_transform(x, transform_matrix, channel_axis=2, fill_mode='nearest', fill_value=0.):
	x = np.rollaxis(x, channel_axis, 0)

ncullen93

"""
Affine transforms implemented on torch tensors, and
only requiring one interpolation

Included:
- Affine()
- AffineCompose()
- Rotation()
- Translation()
- Shear()

ncullen93

"""
Custom datasets from both in-memory and out-of-memory data
"""

import torch.utils.data as data

from PIL import Image
import os
import os.path

ncullen93

def pearsonr(x, y):
    """
    Mimics `scipy.stats.pearsonr`

    Arguments
    ---------
    x : 1D torch.Tensor
    y : 1D torch.Tensor

    Returns

ncullen93

class MultiSampler(Sampler):
    """Samples elements more than once in a single pass through the data.

    This allows the number of samples per epoch to be larger than the number
    of samples itself, which can be useful for data augmentation.
    """
    def __init__(self, nb_samples, desired_samples, shuffle=False):
        self.data_samples = nb_samples
        self.desired_samples = desired_samples
        self.shuffle

ncullen93

import torch as th
from torch.autograd import Variable
from torch.autograd.function import Function

class Symeig(Function):

    def __init__(self, eigenvectors=True, upper=True):
        self.eigenvectors = eigenvectors
        self.upper = upper