grids

Functions

interpolate_unstructured(x, y, z[, resolution, method])

Interpolate unstructured data to a structured grid.

is_structured(x, y[, tol])

Determines whether the x and y points form a structured grid.

needs_cyclic_point(lons)

Module Contents

grids.interpolate_unstructured(x, y, z, resolution=1000, method='linear')

Interpolate unstructured data to a structured grid.

This function takes unstructured (scattered) data points and interpolates them to a structured grid, handling NaN values in z and providing options for different interpolation methods. It creates a regular grid based on the given resolution and interpolates the z-values from the unstructured points onto this grid.

Parameters:

  • x (array_like) – 1D array of x-coordinates.

  • y (array_like) – 1D array of y-coordinates.

  • z (array_like) – 1D array of z-values at each (x, y) point.

  • resolution (int, optional) – The number of points along each axis for the structured grid. Default is 1000.

  • method ({'linear', 'nearest', 'cubic'}, optional) –

    The interpolation method to use. Default is ‘linear’. The methods supported are:

    • ’linear’: Linear interpolation between points.

    • ’nearest’: Nearest-neighbor interpolation.

    • ’cubic’: Cubic interpolation, which may produce smoother results.

Returns:

  • grid_x (ndarray) – 2D array representing the x-coordinates of the structured grid.

  • grid_y (ndarray) – 2D array representing the y-coordinates of the structured grid.

  • grid_z (ndarray) – 2D array of interpolated z-values at the grid points. NaNs may be present in regions where interpolation was not possible (e.g., due to large gaps in the data).

grids.is_structured(x, y, tol=1e-05)

Determines whether the x and y points form a structured grid.

This function checks if the x and y coordinate arrays represent a structured grid, i.e., a grid with consistent spacing between points. The function supports 1D arrays (representing coordinates of a grid) and 2D arrays (representing the actual grid coordinates) of x and y.

Parameters:

  • x (array_like) – A 1D or 2D array of x-coordinates. For example, this can be longitude or the x-coordinate in a Cartesian grid.

  • y (array_like) – A 1D or 2D array of y-coordinates. For example, this can be latitude or the y-coordinate in a Cartesian grid.

  • tol (float, optional) – Tolerance for floating-point comparison to account for numerical precision errors when checking spacing consistency. The default is 1e-5.

Returns:

True if the data represents a structured grid, i.e., the spacing between consecutive points in both x and y is consistent. False otherwise.

Return type:

bool

grids.needs_cyclic_point(lons)