Covariance

These stages compute covariances of measurements

• TXFourierGaussianCovariance - Compute a Gaussian Fourier-space covariance with TJPCov using f_sky only

• TXRealGaussianCovariance - Compute a Gaussian real-space covariance with TJPCov using f_sky only

• TXFourierTJPCovariance - Compute a Gaussian Fourier-space covariance with TJPCov using mask geometry

• TXFourierNamasterCovariance - Compute a Gaussian Fourier-space covariance with NaMaster

• TXRealNamasterCovariance - Compute a Gaussian real-space covariance with NaMaster

class txpipe.covariance.TXFourierGaussianCovariance(*args: Any, **kwargs: Any)

Compute a Gaussian Fourier-space covariance with TJPCov using f_sky only

It imports TJPCov to do so, and runs at a fiducial cosmology.

This version does not account for mask geometry, only the total sky area measured.

class txpipe.covariance.TXRealGaussianCovariance(*args: Any, **kwargs: Any)

Compute a Gaussian real-space covariance with TJPCov using f_sky only

This version does not account for mask geometry, only the total sky area measured.

It is implemented as a subclass of the Fourier-space version, so also uses TJPCov and a fiducial cosmology.

class txpipe.covariance.TXFourierTJPCovariance(*args: Any, **kwargs: Any)

Compute a Gaussian Fourier-space covariance with TJPCov using mask geometry

This also calls out to TJPCov, using more recent additions to that package.

This version, for speed, re-uses the workspace objects cached in the twopoint fourier measurement stage.

class txpipe.covariance_nmt.TXFourierNamasterCovariance(*args: Any, **kwargs: Any)

Compute a Gaussian Fourier-space covariance with NaMaster

This functionality duplicates that of TXFourierTJPCovariance, and we should rationalize.

class txpipe.covariance_nmt.TXRealNamasterCovariance(*args: Any, **kwargs: Any)

Compute a Gaussian real-space covariance with NaMaster

We don’t yet have another stage for this, but should rationalize when comparing to TJPCov.