cars.core.utils

Utils module: contains some cars global shared general purpose functions

Functions

safe_makedirs(directory[, cleanup])	Create directories even if they already exist (mkdir -p)
make_relative_path_absolute(path, directory)	If path is a valid relative path with respect to directory,
safe_cast_float(data)	Safe cast to float data
get_elevation_range_from_metadata(→ Tuple[float, float])	This function will try to derive a valid RPC altitude range
angle_vectors(→ float)	Compute the smallest angle in radians between two angle_vectors

Module Contents

cars.core.utils.safe_makedirs(directory, cleanup=False)

Create directories even if they already exist (mkdir -p)

Parameters:

directory – path of the directory to create

cars.core.utils.make_relative_path_absolute(path, directory)

If path is a valid relative path with respect to directory, returns it as an absolute path

Parameters:

• path (string) – The relative path

• directory (string) – The directory path should be relative to

Returns:

os.path.join(directory,path) if path is a valid relative path form directory, else path

Return type:

string

cars.core.utils.safe_cast_float(data)

Safe cast to float data

Parameters:

data – string to get float in

:return float from data

cars.core.utils.get_elevation_range_from_metadata(img: str, default_min: float = 0, default_max: float = 300) → Tuple[float, float]

This function will try to derive a valid RPC altitude range from img metadata.

It will first try to read metadata with gdal. If it fails, it will look for values in the geom file if it exists If it fails, it will return the default range

Parameters:

• img – Path to the img for which the elevation range is required

• default_min – Default minimum value to return if everything else fails

• default_max – Default minimum value to return if everything else fails

Returns:

(elev_min, elev_max) float tuple

cars.core.utils.angle_vectors(vector_1: numpy.ndarray, vector_2: numpy.ndarray) → float

Compute the smallest angle in radians between two angle_vectors Use arctan2 more precise than arcos2 Tan θ = abs(axb)/ (a.b) (same: Cos θ = (a.b)/(abs(a)abs(b)))

Parameters:

• vector_1 – Numpy first vector

• vector_2 – Numpy second vector

Returns:

Smallest angle in radians