wf_psf.psf_models.tf_modules module

class wf_psf.psf_models.tf_modules.TF_Zernike_mono_PSF(phase_N, lambda_obs, obscurations, zernike_maps, output_dim=64, name=None)[source]

Bases: Module

Build a monochromatic PSF from zernike coefficients.

Following a Zernike model.

Attributes:

name: Returns the name of this module as passed or determined in the ctor.
name_scope: Returns a tf.name_scope instance for this class.
non_trainable_variables: Sequence of non-trainable variables owned by this module and its submodules.
submodules: Sequence of all sub-modules.
trainable_variables: Sequence of trainable variables owned by this module and its submodules.
variables: Sequence of variables owned by this module and its submodules.

Methods

`__call__`(z_coeffs)	Call self as a function.
`with_name_scope`(method)	Decorator to automatically enter the module name scope.

class wf_psf.psf_models.tf_modules.TF_build_phase(phase_N, lambda_obs, obscurations, name=None)[source]

Bases: Module

Build complex phase map from OPD map.

Attributes:

name: Returns the name of this module as passed or determined in the ctor.
name_scope: Returns a tf.name_scope instance for this class.
non_trainable_variables: Sequence of non-trainable variables owned by this module and its submodules.
submodules: Sequence of all sub-modules.
trainable_variables: Sequence of trainable variables owned by this module and its submodules.
variables: Sequence of variables owned by this module and its submodules.

Methods

`__call__`(opd)	Build the phase from the opd.
`apply_obscurations`(phase)	Multiply element-wise with the obscurations.
`opd_to_phase`(opd)	Convert from opd to phase.
`with_name_scope`(method)	Decorator to automatically enter the module name scope.
`zero_padding_diffraction`(no_pad_phase)	Pad with zeros corresponding to the required lambda.

apply_obscurations(phase)[source]: Multiply element-wise with the obscurations.

opd_to_phase(opd)[source]: Convert from opd to phase.

zero_padding_diffraction(no_pad_phase)[source]

Pad with zeros corresponding to the required lambda.

Important: To check the original size of the no_pad_phase variable we have to look in the [1] dimension not the [0] as it is the batch.

class wf_psf.psf_models.tf_modules.TF_fft_diffract(output_dim=64, output_Q=2, name=None)[source]

Bases: Module

Diffract the wavefront into a monochromatic PSF.

Parameters:

output_dim (int) – Dimension of the output square postage stamp
output_Q (int) – Downsampling factor. Must be integer.

Attributes:

name: Returns the name of this module as passed or determined in the ctor.
name_scope: Returns a tf.name_scope instance for this class.
non_trainable_variables: Sequence of non-trainable variables owned by this module and its submodules.
submodules: Sequence of all sub-modules.
trainable_variables: Sequence of trainable variables owned by this module and its submodules.
variables: Sequence of variables owned by this module and its submodules.

Methods

`__call__`(input_phase)	Calculate the normalized PSF from the padded phase array.
`tf_crop_img`(image, output_crop_dim)	Crop images with tf methods.
`with_name_scope`(method)	Decorator to automatically enter the module name scope.

crop_img
normalize_psf

crop_img(image)[source]

normalize_psf(psf)[source]

tf_crop_img(image, output_crop_dim)[source]

Crop images with tf methods.

It handles a batch of 2D images: [batch, width, height]

class wf_psf.psf_models.tf_modules.TF_mono_PSF(phase_N, lambda_obs, obscurations, output_Q, output_dim=64, name=None)[source]

Bases: Module

Calculate a monochromatic PSF from an OPD map.

Attributes:

name: Returns the name of this module as passed or determined in the ctor.
name_scope: Returns a tf.name_scope instance for this class.
non_trainable_variables: Sequence of non-trainable variables owned by this module and its submodules.
submodules: Sequence of all sub-modules.
trainable_variables: Sequence of trainable variables owned by this module and its submodules.
variables: Sequence of variables owned by this module and its submodules.

Methods

`__call__`(opd)	Call self as a function.
`with_name_scope`(method)	Decorator to automatically enter the module name scope.

class wf_psf.psf_models.tf_modules.TF_zernike_OPD(zernike_maps, name=None)[source]

Bases: Module

Turn zernike coefficients into an OPD.

Will use all of the Zernike maps provided. Both the Zernike maps and the Zernike coefficients must be provided.

Parameters:

zernike_maps (Tensor (Num_coeffs, x_dim, y_dim)) –
z_coeffs (Tensor (num_star, num_coeffs, 1, 1)) –

Returns:

opd

Return type:

Tensor (num_star, x_dim, y_dim)

Attributes:

name: Returns the name of this module as passed or determined in the ctor.
name_scope: Returns a tf.name_scope instance for this class.
non_trainable_variables: Sequence of non-trainable variables owned by this module and its submodules.
submodules: Sequence of all sub-modules.
trainable_variables: Sequence of trainable variables owned by this module and its submodules.
variables: Sequence of variables owned by this module and its submodules.

Methods

`__call__`(z_coeffs)	Call self as a function.
`with_name_scope`(method)	Decorator to automatically enter the module name scope.