Segment an IFU cube using emission-line kinematics
Source:R/kinematics_run_manga_bar_model.R
segment_kinematics.RdBuilds native emission-line maps, then clusters either the kinematic feature cube or the path-signature feature cube. It deliberately does not run the ordinary full-spectrum Capivara segmentation: use [segment()] or [segment_large()] when spectral segmentation is the scientific objective.
Usage
segment_kinematics(
cube_path,
redshift = NA_real_,
emission_line = "halpha",
segmentation_mode = c("kinematic", "path_signature"),
output_dir = NULL,
object_id = NULL,
knn_k = 50,
n_segments = 25,
n_path_segments = 45,
starlet_scales = "2:5",
include_coarse_starlet = FALSE,
support_mode = c("starlet", "line_flux"),
line_flux_sigma = 3,
show_plots = interactive()
)Arguments
- cube_path
Path to a MaNGA LOGCUBE.
- redshift
Optional redshift. If `NA`, uses FITS header then local MaNGA metadata.
- emission_line
Emission line alias, such as `"halpha"` or `"oiii5007"`.
- segmentation_mode
`"kinematic"` clusters flux, velocity, dispersion, and profile-shape maps. `"path_signature"` additionally computes the path-signature feature segmentation.
- output_dir
Directory for saved products. Defaults beside the cube.
- object_id
Optional output identifier.
- knn_k
kNN graph size for sparse Ward clustering.
- n_segments
Number of kinematic-aware segments.
- n_path_segments
Number of path-signature segments.
- starlet_scales
Starlet support scales.
- include_coarse_starlet
Include the coarse starlet plane in support.
- support_mode
Foreground support for the kinematic maps. `"starlet"` keeps the white-light starlet footprint; `"line_flux"` trims that footprint with a robust emission-line-flux threshold before fitting.
- line_flux_sigma
Border-noise threshold, in robust sigma units, when `support_mode = "line_flux"`.
- show_plots
Print the compact kinematic panel.