Gridder#

class pyvisgrid.core.gridder.Gridder(u_meter: ndarray, v_meter: ndarray, times: ndarray, img_size: int, fov: float, src_ra: float, src_dec: float, ref_frequency: float, frequency_offsets: _Buffer | _SupportsArray[dtype[Any]] | _NestedSequence[_SupportsArray[dtype[Any]]] | complex | bytes | str | _NestedSequence[complex | bytes | str], antenna_layout: Layout | None = None)[source]#

Bases: object

Methods Summary

from_fits(path, img_size, fov[, uv_colnames])

Initializes the gridder with the visibility data in a given FITS file using the default Gridder for the radionets-project.

from_ms(path, img_size, fov, desc_id[, ...])

Initializes the Gridder with a measurement which is saved in an NRAO CASA Measurement Set.

from_pyvisgen(vis_data, obs, img_size, fov)

Initializes the gridder with the visibility data which is generated by the pyvisgen.simulation.vis_loop function.

from_uvh5(file, *, fov[, stokes_components, ...])

Initializes the gridder with visibility data from the UVH5 file format.

grid([stokes_component])

Grids given visibility data using the default Gridder for the radionets-project.

grid_time_series(step_size[, ...])

Grids the visibilites in a time series so that time steps get gridded individually.

plot_dirty_image([stokes_component])

Plots the (u,v) dirty image, meaning the 2d Fourier transform of the gridded visibilities.

plot_mask([stokes_component])

Plots the (u,v) mask (the binned visibilities) of the gridded interferometric image.

plot_ungridded_uv(**kwargs)

Plots the ungridded (u,v) points as a scatter plot.

Methods Documentation

classmethod from_fits(path: str, img_size: int, fov: float, uv_colnames: dict = None) → GridData[source]#

Initializes the gridder with the visibility data in a given FITS file using the default Gridder for the radionets-project. Currently only extraction of the Stokes I component is supported. More on the Gridder can be found in the constructor of the Gridder.

Parameters:
pathstr

The path to the FITS file.

img_sizeint

The size of the image in pixels.

fovfloat

The physical size of the image in asec.

uv_colnamesdict, optional

Alternative names for the U and V columns in the FITS file. Default is {‘u’: None, ‘v’: None}, meaning the default values of ‘UU’ and ‘VV’ or ‘UU–’ and ‘VV–’ will be used.

classmethod from_ms(path: str, img_size: int, fov: float, desc_id: int, ref_frequency_id: int = 0, use_calibrated: bool = False, filter_flagged: bool = True) → GridData[source]#

Initializes the Gridder with a measurement which is saved in an NRAO CASA Measurement Set. Currently only extraction of the Stokes I component is supported.

Parameters:
path: str

The path of the measurement set root directory.

img_size: int

The size of the image in pixels.

fov: float

The physical size of the image in asec.

desc_id: int

The description id of the visibilites which should be gridded. This id corresponds to a way of choosing between the spectral windows of the observation and its polarization setup.

Note: Currently it is not supported to grid all visibilities of the observation in one image. It is planned to add that at a later point.

ref_frequency_id: int, optional

The index of the reference frequency that will be used if the measurement is a composite observation and no desc_id is given.

use_calibrated: bool, optional

Whether to use the calibrated data from the MODEL_DATA column or the raw data from the DATA column. Default is True.

filter_flagged: bool, optional

Whether to filter out flagged data rows. Default is True.

classmethod from_pyvisgen(vis_data: Visibilities, obs: Observation, img_size: int, fov: float, stokes_components: list[str] | str = 'I', polarizations: list[str] | str | None = None) → GridData[source]#

Initializes the gridder with the visibility data which is generated by the pyvisgen.simulation.vis_loop function. Additionally one can define which stokes components should be calculated with a given polarization.

More on the Gridder can be found in the constructor of the Gridder.

Parameters:
obspyvisgen.simulation.Observation

The observation which is returned by the pyvisgen.simulation.vis_loop function.

vis_datapyvisgen.simulation.Visibilities

The visibility data which is returned by the pyvisgen.simulation.vis_loop function.

img_sizeint

The size of the image in pixels.

fovfloat

The physical size of the image in asec.

stokes_componentslist[str] | str, optional

The Stokes components which are to be calculated and saved in the gridder. This can either be a list of components (e.g. ['I', 'V']) or a single string. Default is 'I'.

polarizationslist[str] | str | None, optional

The polarization type. Default is None.

classmethod from_uvh5(file: str | Path, *, fov: float, stokes_components: str | list[str] = 'I', polarizations: str | list[str] | None = None, station_ids_unavail: float = 0.0, img_size: int | None = None)[source]#

Initializes the gridder with visibility data from the UVH5 file format.

This method allows to select stokes components that will be computed for a given polarization.

Parameters:
filestr or Path

Path to the file that you want to grid. This file has to be in the UVH5 file format from pyvisgen.

fovfloat

The physical size of the image in asec.

stokes_componentsstr | list[str], optional

The Stokes components which are to be calculated and saved in the gridder. This can either be a list of components (e.g. ['I', 'V']) or a single string. Default: 'I'.

polarizationsstr | list[str] | None, optional

The polarization type. Default: None.

station_ids_unavailfloat, optional

Percentage of station ids that are unavailable during the observation. This allows simulating reduced operation capacity even after the main simulation in pyvisgen and thus avoids having to rerun the same simulation with different reduced arrays. Default: 0.0

img_sizeint or None, optional

Image size is read from the file. In case the image size cannot be read, e.g. because the sky image was not saved to the file (“file/sky/SI”), or you want to grid the image to a different size, it can be set through this keyword argument. Default: None

See also

UVH5Writer

The UVH5 data writer class as implemented in pyvisgen, outlining the file structure.

Examples

>>> gridded_vis = Gridder.from_uvh5(
...     "/path/to/example.uvh5",
...     fov=0.024,
... ).grid()
>>> gridded_vis.vis_data.shape
(1344,)

Setting the station_ids_unavail keyword argument to a value greater 0.0, you can reduce the array, without the need to re-simulate the entire dataset:

>>> gridded_vis_reduced = Gridder.from_uvh5(
...     "/path/to/example.uvh5",
...     fov=0.024,
...     station_ids_unavail=0.5,
... ).grid()
>>> gridded_vis_reduced.vis_data.shape
(288,)

Note that this only disables a randomized selection of antennas for the entire observation. Depending on the visibility of the source, the data taken by the antennas may vary during measurement, and the resulting visibilites dataset may be smaller or larger than expected (see, e.g., the shapes (1344,) vs (288,) although only half of the antennas were unavailable).

grid(stokes_component: str = 'I') → GridData[source]#

Grids given visibility data using the default Gridder for the radionets-project.

Parameters:
stokes_componentstr, optional

The stokes component which should be gridded. The specified component has to be initialized first! Otherwise this will result in a KeyError. Default is 'I'.

Returns:
masknumpy.ndarray, optional

The mask created from the given (u,v) coordinates. The mask contains the number of (u,v) coordinates per pixel.

mask_realnumpy.ndarray, optional

The gridded real part of the visibilites.

mask_imagnumpy.ndarray, optional

The gridded imaginary part of the visibilities.

dirty_img_complexnumpy.ndarray, optional

The complex Dirty Image. This is the 2-dimensional Fourier transform of the gridded visibilities.

grid_time_series(step_size: int, time_start_idx: int = 0, time_end_idx: int | None = None, stokes_component: str = 'I', show_progress: bool = True) → GridDataSeries[source]#

Grids the visibilites in a time series so that time steps get gridded individually.

Parameters:
step_sizeint

The number of visibilites (time steps) per gridded time step. Meaning that the resulting series will have len(Gridder) // step_size for its length.

time_start_idxint, optional

The time index at which the time series starts. Default is 0, meaning the time series starts at the first visibility.

time_endint | None, optional

The time index at which the time series end. Default is None, meaning that the time series ends at the last visibility.

stokes_componentstr, optional

The Stokes component to grid. Default is 'I'.

show_progressbool, optional

Whether to show a progress bar for the gridding progress. Default is True.

Returns:
GridDataSeries

The series of gridded and ungridded visibilities.

plot_dirty_image(stokes_component: str = 'I', **kwargs) → tuple[matplotlib.figure.Figure, matplotlib.axes.Axes][source]#

Plots the (u,v) dirty image, meaning the 2d Fourier transform of the gridded visibilities.

Parameters:
stokes_componentstr, optional

The symbol of the stokes component whose dirty image should be plotted. The specified component has to be initialized and gridded first! Otherwise this will result in a KeyError. Default is 'I'.

modestr, optional

The mode specifying which values of the dirty image should be plotted. Possible values are:

  • real: Plots the real part of the dirty image.

  • imag: Plots the imaginary part of the dirty image.

  • abs: Plot the absolute value of the dirty image.

Default is real.

ax_unit: str | astropy.units.Unit, optional

The unit in which to show the ticks of the x and y-axes in. The y-axis is the Declination (DEC) and the x-axis is the Right Ascension (RA). The latter one is defined as increasing from left to right! The unit has to be given as a string or an astropy.units.Unit. The string must correspond to the string representation of an astropy.units.Unit.

Valid units are either pixel or angle units like arcsec, degree etc. Default is pixel.

center_pos: tuple | None, optional

The coordinate center of the image. The coordinates have to be given in the unit defined in the parameter ax_unit above. If ax_unit is set to pixel this parameter is ignored. Default is None, meaning the coordinates of the axes will be given as relative.

normstr | matplotlib.colors.Normalize | None, optional

The name of the norm or a matplotlib norm. Possible values are:

  • log: Returns a logarithmic norm with clipping on (!), meaning

    values above the maximum will be mapped to the maximum and values below the minimum will be mapped to the minimum, thus avoiding the appearance of a colormaps ‘over’ and ‘under’ colors (e.g. in case of negative values). Depending on the use case this is desirable but in case that it is not, one can set the norm to log_noclip or provide a custom norm.

  • log_noclip: Returns a logarithmic norm with clipping off.

  • centered: Returns a linear norm which centered around zero.

  • sqrt: Returns a power norm with exponent 0.5, meaning the

    square-root of the values.

  • other: A value not declared above will be returned as is,

    meaning that this could be any value which exists in matplotlib itself.

Default is None, meaning no norm will be applied.

colorbar_shrink: float, optional

The shrink parameter of the colorbar. This can be needed if the plot is included as a subplot to adjust the size of the colorbar. Default is 1, meaning original scale.

cmap: str | matplotlib.colors.Colormap, optional

The colormap to be used for the plot. Default is 'inferno'.

plot_argsdict | None, optional

The additional arguments passed to the scatter plot. Default is None.

fig_argsdict | None, optional

The additional arguments passed to the figure. If a figure object is given in the fig parameter, this value will be discarded. Default is None.

save_tostr | PathLike | None, optional

The name of the file to save the plot to. Default is None, meaning the plot won’t be saved.

save_argsdict | None, optional

The additional arguments passed to the fig.savefig call. Default is {"bbox_inches":"tight"}.

figmatplotlib.figure.Figure | None, optional

A custom figure object. If set to None, the ax parameter also has to be None! Default is None.

axmatplotlib.axes.Axes | None, optional

A custom axes object. If set to None, the fig parameter also has to be None! Default is None.

Returns:
figmatplotlib.figure.Figure

The figure object.

axmatplotlib.axes.Axes

The axes object.

plot_mask(stokes_component: str = 'I', **kwargs) → tuple[matplotlib.figure.Figure, matplotlib.axes.Axes][source]#

Plots the (u,v) mask (the binned visibilities) of the gridded interferometric image.

Parameters:
stokes_componentstr, optional

The symbol of the stokes component whose mask should be plotted. The specified component has to be initialized and gridded first! Otherwise this will result in a KeyError. Default is 'I'.

modestr, optional

The mode specifying which values of the mask should be plotted. Possible values are:

  • hist: Plots the number of (u,v) points which are sorted in

    each pixel of the image in the (u,v) space.

  • abs: Plots the absolute value of the gridded visibilities,

    meaning the magnitude of the complex numbers in Euler representation.

  • phase: Plots the phase angle of the gridded visibilities,

    meaning the angle in the exponent of the complex numbers in Euler representation.

  • real: Plots the real part of the gridded visibilities.

  • imag: Plots the imaginary part of the gridded visibilities.

Default is hist.

croptuple[list[float | None]], optional

The crop of the image. This has to have the format ([x_left, x_right], [y_left, y_right]), where the left and right values for each axis are the upper and lower limits of the axes which should be shown. IMPORTANT: If one supplies the plt.imshow an extent parameter via the plot_args parameter, this will be the scale in which one has to give the crop! If not, the crop has to be in pixels.

normstr | matplotlib.colors.Normalize | None, optional

The name of the norm or a matplotlib norm. Possible values are:

  • log: Returns a logarithmic norm with clipping on (!), meaning

    values above the maximum will be mapped to the maximum and values below the minimum will be mapped to the minimum, thus avoiding the appearance of a colormaps ‘over’ and ‘under’ colors (e.g. in case of negative values). Depending on the use case this is desirable but in case that it is not, one can set the norm to log_noclip or provide a custom norm.

  • log_noclip: Returns a logarithmic norm with clipping off.

  • centered: Returns a linear norm which centered around zero.

  • sqrt: Returns a power norm with exponent 0.5, meaning the

    square-root of the values.

  • other: A value not declared above will be returned as is,

    meaning that this could be any value which exists in matplotlib itself.

Default is None, meaning no norm will be applied.

cmap: str | matplotlib.colors.Colormap | None, optional

The colormap to be used for the plot. Default is None, meaning the colormap will be default to a value fitting for the chosen mode.

plot_argsdict, optional

The additional arguments passed to the scatter plot. Default is {"color":"royalblue"}.

fig_argsdict | None, optional

The additional arguments passed to the figure. If a figure object is given in the fig parameter, this value will be discarded. Default is None.

save_tostr | PathLike | None, optional

The name of the file to save the plot to. Default is None, meaning the plot won’t be saved.

save_argsdict | None, optional

The additional arguments passed to the fig.savefig call. Default is {"bbox_inches":"tight"}.

figmatplotlib.figure.Figure | None, optional

A custom figure object. If set to None, the ax parameter also has to be None! Default is None.

axmatplotlib.axes.Axes | None, optional

A custom axes object. If set to None, the fig parameter also has to be None! Default is None.

Returns:
figmatplotlib.figure.Figure

The figure object.

axmatplotlib.axes.Axes

The axes object.

plot_ungridded_uv(**kwargs) → tuple[matplotlib.figure.Figure, matplotlib.axes.Axes][source]#

Plots the ungridded (u,v) points as a scatter plot.

Parameters:
modestr, optional

The mode specifying the scale of the (u,v) coordinates. This can be either wave, meaning the coordinates are plotted in units of the reference wavelength, or meter, meaning the (u,v) coordinates will be plotted in meter. Default is wave.

show_timesbool, optional

Whether to show the timestamps of the measured visibilities as a colormap. Default is True.

use_relative_timebool, optional

Whether to show the times relative to the timestamp of the first measurement in hours. Default is True.

times_cmap: str | matplotlib.colors.Colormap, optional

The colormap to be used for the time component of the plot. Default is 'inferno'.

marker_sizefloat | None, optional

The size of the scatter markers in points**2. Default is None, meaning the default value supplied by your matplotlib rcParams.

plot_argsdict, optional

The additional arguments passed to the scatter plot. Default is {"color":"royalblue"}.

fig_argsdict, optional

The additional arguments passed to the figure. If a figure object is given in the fig parameter, this value will be discarded. Default is {}.

save_tostr | None, optional

The name of the file to save the plot to. Default is None, meaning the plot won’t be saved.

save_argsdict, optional

The additional arguments passed to the fig.savefig call. Default is {"bbox_inches":"tight"}.

figmatplotlib.figure.Figure | None, optional

A custom figure object. If set to None, the ax parameter also has to be None! Default is None.

axmatplotlib.axes.Axes | None, optional

A custom axes object. If set to None, the fig parameter also has to be None! Default is None.

Returns:
figmatplotlib.figure.Figure

The figure object.

axmatplotlib.axes.Axes

The axes object.