contour

Draw contour lines of a 2D scalar field. Jarvis-PLOT adds a credible-region mode that computes statistically meaningful levels (HPD or profile-likelihood) for you.

Matplotlib: Axes.contour. The x/y/z coordinates are assembled into the Matplotlib X, Y, Z grid.

Axes: rectangular (ax).

Coordinates

Key	Required	Meaning
x	yes	Grid x positions
y	yes	Grid y positions
z	yes	Scalar field to contour

The field must be gridded — feed it from a density/profile transform or from a shared make_interp_2d result. For contours straight from scattered points, use jpcontour.

Style

Plain contours

Key	Type	Purpose
levels	int or list	Number of levels, or explicit level values
colors	color list	One color per level
linewidths	float list	One width per level
linestyles	string/list	"solid", "dashed", …
cmap	string	Color levels by a colormap (instead of colors)
zorder	number	Draw order

Credible-region mode

Set contour_mode to let Jarvis-PLOT compute the levels:

Key	Purpose
contour_mode	posterior_hpd (highest posterior density) or profile_likelihood
masses	For posterior_hpd: credible masses, e.g. [0.6827, 0.9545] (1σ, 2σ)
sigma	For profile_likelihood: levels in σ, e.g. [1, 2]
ndof	For profile_likelihood: chi-square degrees of freedom (default 2)
labels	Optional per-level labels drawn on the contours

Example — 1σ / 2σ HPD contours over a density

- name: hpd
  data:
    - source: density_grid          # a shared posterior-density result
  axes: ax
  method: contour
  coordinates:
    x: {expr: x}
    y: {expr: y}
    z: {expr: density}
  style:
    contour_mode: posterior_hpd
    masses: [0.6827, 0.9545]
    colors: [black, white]
    linewidths: [0.3, 0.3]
    labels: ["$1\\sigma$", "$2\\sigma$"]
    zorder: 20

Notes

• posterior_hpd is for probability densities; profile_likelihood is for likelihood/χ² fields (use sigma + ndof).
• For filled contours use contourf.
• The posterior_2d and profile_2d figure types add these contours automatically.

---

See also: Plot Methods index · contourf · jpcontour