FDWhittleLikelihood§
- class typed_lisa_toolkit.types.FDWhittleLikelihood[source]§
Whittle likelihood for frequency-domain data.
Assuming the noise is additive, stationary and Gaussian. Note \(d\) the data, \(h\) the template, the log-likelihood is given by
\[\log \mathcal{L} = -\frac{1}{2} \left( d - h \middle| d - h \right).\]We can rewrite this as
\[\log \mathcal{L} = -\frac{1}{2} \left( d \middle| d \right) + \left( d \middle| h \right) -\frac{1}{2} \left( h \middle| h \right).\]The term \(\left( d \middle| d \right)\) is computed upon initialization and is constant.
Attention
This class is considered experimental. If you are interested in using it, please reach out to the developers to discuss your use case and how we can best support it.
- classmethod log_likelihood(log_likelihood_ratio: Array[Any, ModuleType] | ndarray[tuple[Any, ...], dtype[Any]] | complex, data_square: Array[Any, ModuleType] | ndarray[tuple[Any, ...], dtype[Any]] | complex) Array[Any, ModuleType] | ndarray[tuple[Any, ...], dtype[Any]]§
Compute the log likelihood.
- classmethod log_likelihood_ratio(cross_product: Array[Any, ModuleType] | ndarray[tuple[Any, ...], dtype[Any]] | complex, template_square: Array[Any, ModuleType] | ndarray[tuple[Any, ...], dtype[Any]] | complex) Array[Any, ModuleType] | ndarray[tuple[Any, ...], dtype[Any]]§
Compute the log likelihood ratio.
- get_cross_product(template: _FDUniformHomogeneous[Harmonic])[source]§
Get the cross product.
This method computes the term \(\left( d \middle| h \right)\).
- get_log_likelihood_ratio(template: _FDUniformHomogeneous[Harmonic])[source]§
Get the log likelihood ratio.
- get_template_square(template: _FDUniformHomogeneous[Harmonic])[source]§
Get the template square.
This method computes the term \(\left( h \middle| h \right)\).
- property data§
Get the data.
- property noisemodel: NoiseModelT§
Get the noise model.