Geometry

Surface geometry computations from grid data.

Functions

`curvature`

from lst_tools import curvature

kappa = curvature(grid)

Compute the surface curvature distribution.

lst_tools.geometry.curvature

Wall curvature computation along a grid i-line.

curvature

curvature(grid: Grid, *, j: int = 0, smooth: bool = True, debug_path: Path | str | None = None, method: str = 'spline', method_params: dict[str, Any] | None = None) -> np.ndarray

Compute wall curvature along an i-line.

Treats the wall as y(x) and computes: kappa = -y'' / (1 + (y')^2)^(3/2)

Parameters:

Name	Type	Description	Default
`grid`	`Grid`	Structured grid with shape (ny, nx).	required
`j`	`int`	Row index (0..ny-1) along which to compute curvature.	`0`
`smooth`	`bool`	If True, apply smoothing to the raw curvature.	`True`
`debug_path`	`Path \| str \| None`	Directory for debug output (Tecplot ASCII).	`None`
`method`	`str`	Smoothing method ("spline", "savgol", "gaussian", "robust").	`'spline'`
`method_params`	`dict[str, Any] \| None`	Extra keyword arguments passed to the smoother.	`None`

Returns:

Type	Description
`ndarray`	1-D array of curvature values along the i-line.

smooth_gaussian

smooth_gaussian(kappa: ndarray, sigma_frac: float = 0.02) -> np.ndarray

Gaussian smoothing with moving-average fallback.

Parameters:

Name	Type	Description	Default
`kappa`	`ndarray`	Raw curvature array.	required
`sigma_frac`	`float`	Gaussian sigma as a fraction of array length.	`0.02`

Returns:

Type	Description
`ndarray`	Smoothed curvature array.

smooth_kappa

smooth_kappa(x: ndarray, kappa: ndarray, *, method: str = 'spline', method_params: dict[str, Any] | None = None, window_frac: float | None = None, polyorder: int | None = None, sigma_frac: float | None = None, s_factor: float | None = None, median_frac: float | None = None, gauss_frac: float | None = None) -> np.ndarray

Dispatch to a chosen smoothing method.

Parameters:

Name	Type	Description	Default
`x`	`ndarray`	Coordinate array.	required
`kappa`	`ndarray`	Raw curvature array.	required
`method`	`str`	One of "spline", "savgol", "gaussian", "robust".	`'spline'`
`method_params`	`dict[str, Any] \| None`	Extra parameters forwarded to the chosen smoother.	`None`
`window_frac`	`float \| None`	Optional Savitzky-Golay window fraction.	`None`
`polyorder`	`int \| None`	Optional Savitzky-Golay polynomial order.	`None`
`sigma_frac`	`float \| None`	Optional Gaussian sigma fraction.	`None`
`s_factor`	`float \| None`	Optional spline smoothing factor.	`None`
`median_frac`	`float \| None`	Optional robust median-filter window fraction.	`None`
`gauss_frac`	`float \| None`	Optional robust Gaussian smoothing fraction.	`None`

Returns:

Type	Description
`ndarray`	Smoothed curvature array.

smooth_robust

smooth_robust(kappa: ndarray, median_frac: float = 0.01, gauss_frac: float = 0.02) -> np.ndarray

Median filter to remove spikes, then Gaussian smooth.

Parameters:

Name	Type	Description	Default
`kappa`	`ndarray`	Raw curvature array.	required
`median_frac`	`float`	Median filter window as a fraction of array length.	`0.01`
`gauss_frac`	`float`	Gaussian sigma fraction passed to `smooth_gaussian`.	`0.02`

Returns:

Type	Description
`ndarray`	Smoothed curvature array.

smooth_savgol

smooth_savgol(kappa: ndarray, window_frac: float = 0.03, polyorder: int = 3) -> np.ndarray

Savitzky-Golay smoothing with moving-average fallback.

Parameters:

Name	Type	Description	Default
`kappa`	`ndarray`	Raw curvature array.	required
`window_frac`	`float`	Window size as a fraction of array length (forced odd).	`0.03`
`polyorder`	`int`	Polynomial order (typically 2-3).	`3`

Returns:

Type	Description
`ndarray`	Smoothed curvature array.

smooth_spline

smooth_spline(x: ndarray, kappa: ndarray, s_factor: float = 0.01) -> np.ndarray

Smoothing spline on non-uniform x with moving-average fallback.

The smoothing parameter s is derived from a median-based noise estimate scaled by s_factor.

Parameters:

Name	Type	Description	Default
`x`	`ndarray`	Coordinate array (non-uniform spacing OK).	required
`kappa`	`ndarray`	Raw curvature array.	required
`s_factor`	`float`	Scale factor for the smoothing parameter.	`0.01`

Returns:

Type	Description
`ndarray`	Smoothed curvature array.

`curvilinear_coordinate`

from lst_tools import curvilinear_coordinate

s = curvilinear_coordinate(grid)

Compute the curvilinear coordinate along the surface.

lst_tools.geometry.curvilinear_coordinate

Curvilinear (streamwise) coordinate computation along a grid line.

curvilinear_coordinate

curvilinear_coordinate(x: Any, y: Any, j: int = 0, debug_path: Path | str | None = None) -> np.ndarray

Return a curvilinear coordinate array s for a given grid line.

This minimal version mirrors the original Fortran logic::

1	`s(i) = x(i, 1)`

i.e., use the x-coordinates along a fixed j-line as the streamwise coordinate.

Parameters:

Name	Type	Description	Default
`x`	`Any`	2-D grid x-coordinates with shape (ny, nx), or 1-D array.	required
`y`	`Any`	2-D grid y-coordinates with shape (ny, nx), or 1-D array.	required
`j`	`int`	Row index (0..ny-1) along which to extract s.	`0`
`debug_path`	`Path \| str \| None`	Directory for Tecplot ASCII debug output.	`None`

Returns:

Type	Description
`ndarray`	1-D array of the curvilinear coordinate along the selected j-line.

Note

If 1-D arrays are provided, the function simply returns a copy of x. This is intentionally simple. If you later want arc-length or more sophisticated definitions of s, extend this routine while keeping the same call signature.

`surface_angle`

from lst_tools import surface_angle

theta = surface_angle(grid)

Compute the surface angle distribution.

lst_tools.geometry.surface_angle

Surface-angle computation along a wall j-line.

surface_angle

surface_angle(grid: Grid, *, method: str = 'first_order', debug_path: Path | str | None = None) -> np.ndarray

Compute surface tangent angle phi along the wall j-line.

Parameters:

Name	Type	Description	Default
`grid`	`Grid`	Structured grid (shape ny x nx).	required
`method`	`str`	Differentiation scheme — `"first_order"` (forward/backward differences) or `"second_order"` (`np.gradient`).	`'first_order'`
`debug_path`	`Path \| str \| None`	Directory for Tecplot ASCII debug output.	`None`

Returns:

Type	Description
`ndarray`	Surface tangent angle phi (radians) at each streamwise station.

Raises:

Type	Description
`ValueError`	If method is not a recognised scheme.

`radius`

from lst_tools import radius

r = radius(grid)

Compute the local body radius.

lst_tools.geometry.radius

Local body radius computation for different geometry kinds.

radius

radius(grid: Grid, cfg: dict, *, debug_path: Path | str | None = None) -> np.ndarray

Compute the local body radius array for the given geometry kind.

Parameters:

Name	Type	Description	Default
`grid`	`Grid`	Computational grid with shape (ny, nx).	required
`cfg`	`dict`	Configuration object with `geometry.type` and related params.	required
`debug_path`	`Path \| str \| None`	Directory for Tecplot ASCII debug output.	`None`

Returns:

Type	Description
`ndarray`	1-D radius array along the wall (j = 0).

Geometry Kinds

`GeometryKind`

from lst_tools import GeometryKind, list_geometry_kinds

kinds = list_geometry_kinds()

Enumeration of supported geometry types.

lst_tools.geometry.GeometryKind

Bases: IntEnum

Integer identifiers for supported geometry types.

lst_tools.geometry.list_geometry_kinds

list_geometry_kinds() -> Mapping[int, str]

Return a mapping {integer_id: 'NAME — explanation'} for display/CLI.