linna.nn

Module Contents

Classes

ResBlock_batchnorm	Residual block
ChtoModelv2	main neural network used by linna
ChtoModelv2_linear	For testing
LinearModel	Linear regression model
pytorchPolynomialLinear	Implement a polynomial fit using pytorch

class linna.nn.ResBlock_batchnorm(in_size, channel, out_size)

Bases: torch.nn.Module

Residual block

init_weight(self)

initialize the weight of neural network

forward(self, x)

Parameters

s (torch tensor) – input array

Returns

ourput array

Return type

torch tensor

class linna.nn.ChtoModelv2(in_size, out_size, linearmodel, docpu=False)

Bases: torch.nn.Module

main neural network used by linna

init_weight(self)

initialize weights for neural network

forward(self, s)

Parameters

s (torch tensor) – input array

Returns

ourput array

Return type

torch tensor

class linna.nn.ChtoModelv2_linear(in_size, out_size, linearmodel, docpu=False)

Bases: torch.nn.Module

For testing

init_weight(self)

forward(self, s)

class linna.nn.LinearModel(norder, npc, x_transform=None, y_transform=None, y_inverse_transform_data=None)

Linear regression model

train(self, train_x, train_y, sample_weight=None)

__call__(self, x)

istrained(self)

save(self, outname)

predict(self, x)

class linna.nn.pytorchPolynomialLinear(ndegree)

Implement a polynomial fit using pytorch

fit(self, train_x, train_y, sample_weight=None)

trainsform(self, tensor)

__call__(self, X)