rm_lite.utils.multiscale¶

RM-ms-clean utils

Attributes¶

`KERNEL_FUNCS`
`MSG`
`TQDM_OUT`

Functions¶

`_scale_bias_function`(→ float)	Offringa et al. (2017) scale-bias function.
`convolve_fdf_scale`(→ numpy.typing.NDArray[numpy.float64])	Convolve the FDF with a Gaussian kernel.
`find_significant_scale`(→ tuple[float, float])
`gaussian_scale_kernel_function`(...)	Gaussian scale kernel function.
`hanning`(→ numpy.typing.NDArray[numpy.float64])	Hanning window function.
`multiscale_cycle`(→ rm_lite.utils.clean.CleanLoopResults)
`multiscale_minor_loop`(...)
`mutliscale_rmclean`(→ rm_lite.utils.clean.RMCleanResults)
`scale_bias_function`(→ numpy.typing.NDArray[numpy.float64])	Offringa et al. (2017) scale-bias function.
`scale_bias_function_cornwell`(...)
`tapered_quad_kernel_function`(...)	Tapered quadratic kernel function.

Module Contents¶

rm_lite.utils.multiscale._scale_bias_function(scale: float, scale_0: float, scale_bias: float) → float¶

Offringa et al. (2017) scale-bias function.

Parameters:

scale (float) – Scale parameter (relative to PSF FWHM)
scale_0 (float) – The first non-zero scale parameter
scale_bias (float) – The scale-bias parameter

Returns:

Weighting factor per scale

Return type:

float

rm_lite.utils.multiscale.convolve_fdf_scale(scale: float, fwhm: float, fdf_arr: numpy.typing.NDArray[numpy.float64], phi_double_arr_radm2: numpy.typing.NDArray[numpy.float64], kernel: Literal['tapered_quad', 'gaussian'] = 'gaussian', sum_normalised: bool = True) → numpy.typing.NDArray[numpy.float64]¶

Convolve the FDF with a Gaussian kernel.

Parameters:

scale (float) – Scale parameter (relative to PSF FWHM)
fwhm (float) – FWHM of the RMSF
fdf_arr (NDArray[np.float64]) – FDF array (complex)
phi_double_arr_radm2 (NDArray[np.float64]) – Double-length Faraday depth array (rad/m^2)
kernel (Literal["tapered_quad", "gaussian"]) – Kernel function

Raises:

ValueError – If an invalid normalization method is provided

Returns:

Convolved FDF array

Return type:

NDArray[np.float64]

rm_lite.utils.multiscale.find_significant_scale(scales: numpy.typing.NDArray[numpy.float64], scale_bias: float, fdf_arr: numpy.typing.NDArray[numpy.float64], phi_double_arr_radm2: numpy.typing.NDArray[numpy.float64], fwhm: float, kernel: Literal['tapered_quad', 'gaussian'] = 'gaussian') → tuple[float, float]¶

rm_lite.utils.multiscale.gaussian_scale_kernel_function(phi_double_arr_radm2: numpy.typing.NDArray[numpy.float64], scale: float, rmsf_fwhm: float, sum_normalised: bool = True) → numpy.typing.NDArray[numpy.float64]¶

Gaussian scale kernel function.

Parameters:

phi_double_arr_radm2 (NDArray[np.float64]) – Phi array in rad/m^2
scale (float) – Scale (in FWHM units)
rmsf_fwhm (float) – RMSF FWHM in rad/m^2

Returns:

Kernel function array (sum normalised)

Return type:

NDArray[np.float64]

rm_lite.utils.multiscale.hanning(x_arr: numpy.typing.NDArray[numpy.float64], length: float) → numpy.typing.NDArray[numpy.float64]¶

Hanning window function.

Parameters:

x_arr (NDArray[np.float64]) – Array of x values
length (float) – Length of the window

Returns:

Hanning window function array

Return type:

NDArray[np.float64]

rm_lite.utils.multiscale.multiscale_cycle(scales: numpy.typing.NDArray[numpy.float64], scale_bias: float, phi_arr_radm2: numpy.typing.NDArray[numpy.float64], phi_double_arr_radm2: numpy.typing.NDArray[numpy.float64], dirty_fdf_spectrum: numpy.typing.NDArray[numpy.float64], rmsf_spectrum: numpy.typing.NDArray[numpy.float64], rmsf_fwhm: float, mask: float, threshold: float, max_iter: int, max_iter_sub_minor: int, gain: float, kernel: Literal['tapered_quad', 'gaussian'] = 'gaussian') → rm_lite.utils.clean.CleanLoopResults¶

rm_lite.utils.multiscale.multiscale_minor_loop(scales: numpy.typing.NDArray[numpy.float64], scale_bias: float, resid_fdf_spectrum_mask: numpy.ma.MaskedArray, phi_arr_radm2: numpy.typing.NDArray[numpy.float64], phi_double_arr_radm2: numpy.typing.NDArray[numpy.float64], rmsf_spectrum: numpy.typing.NDArray[numpy.float64], rmsf_fwhm: float, max_iter: int, max_iter_sub_minor: int, gain: float, mask: float, threshold: float, start_iter: int = 0, update_mask: bool = True, kernel: Literal['tapered_quad', 'gaussian'] = 'gaussian') → rm_lite.utils.clean.MinorLoopResults¶

rm_lite.utils.multiscale.mutliscale_rmclean(freq_arr_hz: numpy.typing.NDArray[numpy.float64], dirty_fdf_arr: numpy.typing.NDArray[numpy.float64], phi_arr_radm2: numpy.typing.NDArray[numpy.float64], rmsf_arr: numpy.typing.NDArray[numpy.float64], phi_double_arr_radm2: numpy.typing.NDArray[numpy.float64], fwhm_rmsf_arr: numpy.typing.NDArray[numpy.float64], mask: float, threshold: float, max_iter: int = 1000, max_iter_sub_minor: int = 10000, gain: float = 0.1, scale_bias: float = 0.9, scales: numpy.typing.NDArray[numpy.float64] | None = None, mask_arr: numpy.typing.NDArray[numpy.float64] | None = None, kernel: Literal['tapered_quad', 'gaussian'] = 'gaussian') → rm_lite.utils.clean.RMCleanResults¶

rm_lite.utils.multiscale.scale_bias_function(scales: numpy.typing.NDArray[numpy.float64], scale_bias: float) → numpy.typing.NDArray[numpy.float64]¶

Offringa et al. (2017) scale-bias function.

Parameters:

scales (NDArray[np.float64]) – Scale parameters (relative to PSF FWHM)
scale_bias (float) – The scale-bias parameter

Returns:

Weighting factors per scale

Return type:

NDArray[np.float64]

rm_lite.utils.multiscale.scale_bias_function_cornwell(scales: numpy.typing.NDArray[numpy.float64]) → numpy.typing.NDArray[numpy.float64]¶

rm_lite.utils.multiscale.tapered_quad_kernel_function(phi_double_arr_radm2: numpy.typing.NDArray[numpy.float64], scale: float, rmsf_fwhm: float, sum_normalised: bool = True) → numpy.typing.NDArray[numpy.float64]¶

Tapered quadratic kernel function.

Parameters:

phi_double_arr_radm2 (NDArray[np.float64]) – Phi array in rad/m^2
scale (float) – Scale (in FWHM units)
rmsf_fwhm (float) – RMSF FWHM in rad/m^2

Returns:

Kernel function array (sum normalised)

Return type:

NDArray[np.float64]

rm_lite.utils.multiscale.KERNEL_FUNCS: dict[str, Callable]¶

rm_lite.utils.multiscale.MSG = 'This module is not yet implemented.'¶

rm_lite.utils.multiscale.TQDM_OUT¶