Struct FourierNetworkCodec

pub struct FourierNetworkCodec {
    pub fourier_features: NonZeroUsize,
    pub fourier_scale: Positive<f64>,
    pub num_blocks: NonZeroUsize,
    pub learning_rate: Positive<f64>,
    pub num_epochs: usize,
    pub mini_batch_size: Option<NonZeroUsize>,
    pub seed: u64,
}

Fourier network codec which trains and overfits a fourier feature neural network on encoding and predicts during decoding.

The approach is based on the papers by Tancik et al. 2020 (https://dl.acm.org/doi/abs/10.5555/3495724.3496356) and by Huang and Hoefler 2020 (https://arxiv.org/abs/2210.12538).

Fields

fourier_features: NonZeroUsize

The number of Fourier features that the data coordinates are projected to

fourier_scale: Positive<f64>

The standard deviation of the Fourier features

num_blocks: NonZeroUsize

The number of blocks in the network

learning_rate: Positive<f64>

The learning rate for the Adam optimizer

num_epochs: usize

The number of epochs for which the network is trained

mini_batch_size: Option<NonZeroUsize>

The optional mini-batch size used during training

Setting the mini-batch size to None disables the use of batching, i.e. the network is trained using one large batch that includes the full data.

seed: u64

The seed for the random number generator used during encoding

Trait Implementations

impl Clone for FourierNetworkCodec

fn clone(&self) -> FourierNetworkCodec

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Codec for FourierNetworkCodec

type Error = FourierNetworkCodecError

Error type that may be returned during encodeing and decodeing.

fn encode(&self, data: AnyCowArray<'_>) -> Result<AnyArray, Self::Error>

Encodes the data and returns the result.

fn decode(&self, _encoded: AnyCowArray<'_>) -> Result<AnyArray, Self::Error>

Decodes the encoded data and returns the result.

fn decode_into( &self, encoded: AnyArrayView<'_>, decoded: AnyArrayViewMut<'_>, ) -> Result<(), Self::Error>

Decodes the encoded data and writes the result into the provided decoded output.

impl<'de> Deserialize<'de> for FourierNetworkCodec

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl JsonSchema for FourierNetworkCodec

fn schema_name() -> Cow<'static, str>

The name of the generated JSON Schema.

fn schema_id() -> Cow<'static, str>

Returns a string that uniquely identifies the schema produced by this type.

fn json_schema(generator: &mut SchemaGenerator) -> Schema

Generates a JSON Schema for this type.

fn always_inline_schema() -> bool

Whether JSON Schemas generated for this type should be included directly in parent schemas, rather than being re-used where possible using the $ref keyword.

impl Serialize for FourierNetworkCodec

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StaticCodec for FourierNetworkCodec

const CODEC_ID: &'static str = "fourier-network"

Codec identifier.

type Config<'de> = FourierNetworkCodec

Configuration type, from which the codec can be created infallibly.

fn from_config(config: Self::Config<'_>) -> Self

Instantiate a codec from its configuration.

fn get_config(&self) -> StaticCodecConfig<'_, Self>

Get the configuration for this codec.