The CM_Utils Functions

The pycoxmunk.CM_Utils module contains utility functions used elsewhere in the code. In general, these should not need to be used by the user but they are documented here:

The gauss_leg_quadx function

This function computes the abscissas and weights of the Gauss-Legendre n-point quadrature. This utility function is included primarily for use within the Cox-Munk BRDF calculations, for computing the diffuse terms, but is available for users if needed. It is called with three arguments:

  • n: Integer, the number of points to compute. Higher n is more accurate in downstream computations, but slower.

  • x1: Float, the lower limit for computation. Often 0.

  • x2: Float, the upper limit for computation. Often π/2.

Two variables are returned:

  • x: An array of length n giving the abscissa values.

  • w: An array of length n giving the weights.

The check_and_reshape function

This utility function checks that a given variable is either a float or an array of the same shape as another, reference, array. It is used during processing to ensure that arrays of different sizes are not passed around the code, it a singular array is given as input, it will be converted to the same shape as the reference. It is called with two arguments:

  • arr: An input variable of any type.

  • good_shape: Tuple giving the shape of a reference array.

The function returns:

  • A float if arr is a float

  • An array of shape good_shape with all values equalling arr if arr is a singular array.

  • An unmodified copy of arr if it is an array of the same shape as good_shape.

A ValueError is raised if arr is neither a float nor an array, or if arr does not have a compatible shape.

The check_type function

This function ensures that a given input variable is of the correct type. It expects the input to be: Any float, a numpy ndarray, an xarray DataArray or a dask array. Numpy and xarray types are converted into dask arrays. Any other type of input raises a TypeError.

Inputs:

  • in_val: The input variable of any type.

  • var_typ: The name of the input variable, solar_zenith_angle for example. Used only when raising TypeError.

Returns:

  • If in_val is a float or dask array, an unmodified in_val is returned.

  • If in_val is a numpy array or xarray, a dask version of in_val is returned.

The _write_gdal function

This function is intended to be private and is used during testing to save intermediate output as TIFF images. It takes two arguments, fname, a string specifying the filename to save an image to, and datas, a 2d array containing the data to be saved.

The _gauss_leg_point function

This function is not intended to be called by users, it is called internally by gauss_leg_quadx.