← Back to API reference

neuraltrain.models.common.ChannelMerger

pydantic model neuraltrain.models.common.ChannelMerger

Configuration for the ChannelMerger module.

Parameters:

• embed_ref – Also embed the reference position, e.g. to enable handling bipolar channels. This requires passing both positions and ref_positions to forward().

• dropout_around_channel – If True, randomly sample a channel to apply dropout around. If False, randomly sample a point in [0, 1] ^ D, where D is the number of dimensions (2 or 3), around which to apply dropout.

• unmerge – If True, unmerge (rather than merge) channels. This is useful to compute the inverse operation of a default ChannelMerger. In this case, the input to forward() should be of shape (B, n_virtual_channels, T).

• invalid_value –

  If all position dimensions for a channel are equal to invalid_value, the channel will be masked out. This is useful when examples within a batch contain different channels and therefore some channels need to be ignored for some of the examples.

  NOTE: ns.extractors.ChannelPositions defines this value as well.

Fields:

• dropout (float)

• dropout_around_channel (bool)

• embed_ref (bool)

• fourier_emb_config (neuraltrain.models.common.FourierEmb)

• invalid_value (float)

• n_subjects (int)

• n_virtual_channels (int)

• per_subject (bool)

• unmerge (bool)

• usage_penalty (float)

field n_virtual_channels: int = 270

field fourier_emb_config: FourierEmb = FourierEmb(n_freqs=None, total_dim=288, n_dims=2, margin=0.2)

field dropout: float = 0

field dropout_around_channel: bool = False

field usage_penalty: float = 0.0

field n_subjects: int = 200

field per_subject: bool = False

field embed_ref: bool = False

field unmerge: bool = False

field invalid_value: float = -0.1

build() → ChannelMergerModel

← Back to API reference