foscat.backend#

Classes#

foscat_backend

Module Contents#

class foscat.backend.foscat_backend(name, mpi_rank=0, all_type='float64', gpupos=0, silent=False)[source]#

TENSORFLOW = 1#

TORCH = 2#

NUMPY = 3#

BACKEND#

float64#

float32#

int64#

int32#

complex64#

complex128#

gpulist#

ngpu = 1#

tf_loc_function(func)[source]#

calc_iso_orient(norient)[source]#

calc_fft_orient(norient, nharm, imaginary)[source]#

bk_SparseTensor(indice, w, dense_shape=[])[source]#

bk_stack(list, axis=0)[source]#

bk_sparse_dense_matmul(smat, mat)[source]#

periodic_pad(x, pad_height, pad_width)[source]#

Applies periodic (‘wrap’) padding to a 4D TensorFlow tensor (N, H, W, C).

Args: x (tf.Tensor): Input tensor with shape (batch_size, height, width, channels).

pad_height (tuple): Tuple (top, bottom) defining the vertical padding size. pad_width (tuple): Tuple (left, right) defining the horizontal padding size.

Returns:: tf.Tensor: Tensor with periodic padding applied.

conv2d(x, w, strides=[1, 1, 1, 1], padding='SAME')[source]#

conv1d(x, w, strides=[1, 1, 1], padding='SAME')[source]#

bk_threshold(x, threshold, greater=True)[source]#

bk_maximum(x1, x2)[source]#

bk_device(device_name)[source]#

bk_ones(shape, dtype=None)[source]#

bk_conv1d(x, w)[source]#

bk_flattenR(x)[source]#

bk_flatten(x)[source]#

bk_resize_image(x, shape)[source]#

bk_L1(x)[source]#

bk_square_comp(x)[source]#

bk_reduce_sum(data, axis=None)[source]#

bk_size(data)[source]#

iso_mean(x, use_2D=False)[source]#

fft_ang(x, nharm=1, imaginary=False, use_2D=False)[source]#

constant(data)[source]#

bk_reduce_mean(data, axis=None)[source]#

bk_reduce_min(data, axis=None)[source]#

bk_random_seed(value)[source]#

bk_random_uniform(shape)[source]#

bk_reduce_std(data, axis=None)[source]#

bk_sqrt(data)[source]#

bk_abs(data)[source]#

bk_is_complex(data)[source]#

bk_distcomp(data)[source]#

bk_norm(data)[source]#

bk_square(data)[source]#

bk_log(data)[source]#

bk_matmul(a, b)[source]#

bk_tensor(data)[source]#

bk_shape_tensor(shape)[source]#

bk_complex(real, imag)[source]#

bk_exp(data)[source]#

bk_min(data)[source]#

bk_argmin(data)[source]#

bk_tanh(data)[source]#

bk_max(data)[source]#

bk_argmax(data)[source]#

bk_reshape(data, shape)[source]#

bk_repeat(data, nn, axis=0)[source]#

bk_tile(data, nn, axis=0)[source]#

bk_roll(data, nn, axis=0)[source]#

bk_expand_dims(data, axis=0)[source]#

bk_transpose(data, thelist)[source]#

bk_concat(data, axis=None)[source]#

bk_zeros(shape, dtype=None)[source]#

bk_gather(data, idx)[source]#

bk_reverse(data, axis=0)[source]#

bk_fft(data)[source]#

bk_fftn(data, dim=None)[source]#

bk_ifftn(data, dim=None, norm=None)[source]#

bk_rfft(data)[source]#

bk_irfft(data)[source]#

bk_conjugate(data)[source]#

bk_real(data)[source]#

bk_imag(data)[source]#

bk_relu(x)[source]#

bk_clip_by_value(x, xmin, xmax)[source]#

bk_cast(x)[source]#

bk_variable(x)[source]#

bk_assign(x, y)[source]#

bk_constant(x)[source]#

to_numpy(x)[source]#