quality_control.quicklook

class kpfpipe.quality_control.quicklook.PlotL0(l0_obj, output_dir=None, full_res=False)

Bases: object

Quicklook plots for KPF L0 (raw CCD) data.

Takes a KPF0 object and generates plots of the raw detector images. Pure visualization — no science computation. Amplifier counting and orientation delegate to ImageAssembly so detector-geometry knowledge has a single home.

Parameters:

• l0_obj (KPF0) – KPF0 data object.

• output_dir (str or None) – Directory to save PNG files. None returns the Figure only.

• full_res (bool) – Save full 1:1 pixel-resolution PNGs instead of the default downsampled figure PNGs. Full-resolution files can be tens of MB per chip.

run(which, *, full_res=None)

Generate the requested plot(s) for every chip that has data.

Saves each to output_dir and, in that save-to-disk mode, closes the matplotlib figure so callers don’t accumulate them. When output_dir is None the figures are returned open, so they display when the caller renders them (e.g. interactively in a notebook).

Parameters:

• which (str) – ‘all’ to run every implemented plot, or the name of a single plot method (one of self._PLOT_METHODS).

• full_res (bool or None) – Save native-size PNGs when true. None uses the constructor’s full_res setting.

Returns:

Maps {method_name}_{chip} to its matplotlib.Figure (closed only when saved to output_dir); useful for tests and introspection.

Return type:

dict

Raises:

ValueError – If which is neither ‘all’ nor a known plot method name.

stitched_image(chip, *, full_res=None)

Plot the stitched raw detector image for one CCD.

Replicates v2.12 plot_L0_stitched_image.

Parameters:

• chip (str) – ‘green’ or ‘red’.

• full_res (bool or None) – Save a native-size, one-output-pixel-per-CCD-pixel PNG when true. None uses the constructor’s full_res setting.

Returns:

The stitched-image figure.

Return type:

matplotlib.Figure

class kpfpipe.quality_control.quicklook.PlotL1(l1_obj, output_dir=None, full_res=False)

Bases: object

Quicklook plots for KPF L1 (assembled FFI) data.

Takes a KPF1 object and generates plots of the assembled detector image. Pure visualization — no science computation. Read-noise header mapping is imported from ImageAssembly so there is one source of truth for which FITS keywords hold which amplifier’s read noise.

Parameters:

• l1_obj (KPF1) – KPF1 data object (post-ImageAssembly).

• output_dir (str or None) – Directory to save PNG files. None returns the Figure only.

• full_res (bool) – Save full 1:1 pixel-resolution PNGs instead of the default downsampled figure PNGs. Full-resolution files can be tens of MB per chip.

bias_subtracted(chip)

Image with master bias subtracted.

Matches v2.12 plot_2D_image(subtract_master_bias=True).

column_cut(chip)

Median column profiles at 10th/50th/90th percentiles. Matches v2.12 plot_2D_column_cut.

dark_subtracted(chip)

Image with master dark subtracted, displayed in e-/hr. Matches v2.12 plot_2D_image(subtract_master_dark=True).

histogram(chip)

Flux distribution histogram. Matches v2.12 plot_2D_image_histogram.

image(chip, *, full_res=None)

Plot the assembled FFI for one CCD.

Replicates v2.12 plot_2D_image for the basic science-frame case.

Parameters:

• chip (str) – ‘green’ or ‘red’.

• full_res (bool or None) – Save a native-size, one-output-pixel-per-CCD-pixel PNG when true. None uses the constructor’s full_res setting.

Returns:

The assembled-FFI figure.

Return type:

matplotlib.Figure

order_trace_overlay(chip)

2x2 grid with order trace overlaid. Matches v2.12 plot_2D_order_trace2x2. Requires calibration DB lookup for order trace file (not yet in v3).

run(which, *, full_res=None)

Generate the requested plot(s) for every chip that has data.

Saves each to output_dir and, in that save-to-disk mode, closes the matplotlib figure so callers don’t accumulate them. When output_dir is None the figures are returned open, so they display when the caller renders them (e.g. interactively in a notebook).

Parameters:

• which (str) – ‘all’ to run every implemented plot, or the name of a single plot method (one of self._PLOT_METHODS).

• full_res (bool or None) – Save native-size PNGs when true. None uses the constructor’s full_res setting.

Returns:

Maps {method_name}_{chip} to its matplotlib.Figure (closed only when saved to output_dir); useful for tests and introspection.

Return type:

dict

Raises:

ValueError – If which is neither ‘all’ nor a known plot method name.

zoom_3x3(chip)

3x3 grid of zoomed detector regions. Matches v2.12 plot_2D_image_zoom_3x3.

class kpfpipe.quality_control.quicklook.PlotL2(l2_obj, output_dir=None, obs_id=None)

Bases: object

Quicklook plots for KPF L2 (extracted 1D spectra) data.

Parameters:

• l2_obj – KPF2 data object (post-SpectralExtraction; requires WavelengthCalibration to have populated the per-fiber WAVE arrays).

• output_dir – directory to save PNG files. None = return Figure only.

• obs_id – observation ID for titles/filenames. If None, falls back to the l2_obj.obs_id attribute (populated on every construction path).

orderlet_flux_ratios(chip)

Per-order orderlet flux ratios (SCI1/SCI2, SCI3/SCI2, SCI1/SCI3, SKY/SCI2, CAL/SCI2) vs order-center wavelength, one panel each, with the all-order median annotated.

peak_flux(chip)

Per-order peak flux (95th pctile counts) for each fiber and the summed-SCI orderlet.

run(which)

Generate the requested plot(s) for every chip that has data, saving each to output_dir. In that save-to-disk mode the figure is closed so callers don’t accumulate them; when output_dir is None the figures are returned open, so they display when the caller renders them (e.g. interactively in a notebook).

Parameters:

which – ‘all’ to run every implemented plot, or the name of a single plot method (one of self._PLOT_METHODS).

Returns:

dict mapping {method_name}_{chip} to matplotlib.Figure (closed only when saved to output_dir; useful for tests/introspection).

snr_per_order(chip)

Per-order SNR (95th pctile of flux/sqrt(|var|)) for each fiber plus the summed-SCI orderlet, vs order-center wavelength.

spectrum_one_row(chip, fibers=None)

Stacked per-fiber spectrum: one panel per fiber, all orders of that fiber concatenated, x-axis is concatenated wavelength.

spectrum_single_order(chip, order=None)

Overplot the science orderlets (and SKY/CAL) for a single spectral order vs wavelength.

Parameters:

• chip – ‘GREEN’ or ‘RED’.

• order – 1-indexed spectral order. Defaults to a representative order near the middle of the chip.

class kpfpipe.quality_control.quicklook.PlotL4(l4_obj, output_dir=None, obs_id=None)

Bases: object

Quicklook plots for KPF L4 (RVs and CCFs) data.

Parameters:

• l4_obj – KPF4 data object (post-CrossCorrelation + RadialVelocity).

• output_dir – directory to save PNG files. None = return Figure only.

• obs_id – observation ID for titles/filenames. If None, falls back to the l4_obj.obs_id attribute (populated on every construction path).

ccf_grid(chip)

Per-orderlet CCF grid for one chip, matching the v2.12 plot_CCF_grid layout: five panels (SCI1, SCI2, SCI3, CAL, SKY), each order’s CCF normalized and offset, colored by the default cycle and labeled by order index. SCI panels add the combined-RV line + value and per-order delta-RV / weight columns. Returns None if the chip has no CCF data.

run(which)

Generate the requested plot(s) for every chip that has CCF data, saving each to output_dir. In that save-to-disk mode the figure is closed so callers don’t accumulate them; when output_dir is None the figures are returned open, so they display when the caller renders them (e.g. interactively in a notebook).

Parameters:

which – ‘all’ to run every implemented plot, or the name of a single plot method (one of self._PLOT_METHODS).

Returns:

dict mapping {method_name}_{chip} to matplotlib.Figure (closed only when saved to output_dir; useful for tests/introspection).

Submodules

level0	L0 quicklook plots for raw KPF detector images.
level1	L1 quicklook plots for assembled FFI.
level2	L2 quicklook plots for extracted KPF 1D spectra.
level4	L4 quicklook plots for KPF cross-correlation functions (CCFs) and RVs.