Faster-Raster

fasterraster is a fast multi-threaded python library for performing raster operations using openMP and numpy objects complete with simple IO.

• Github repository: https://github.com/asenogles/fasterraster
• PyPI: https://pypi.org/project/fasterraster

Motivation

fasterraster was developed to quickly perform raster operations, enabling self-supervised learning for raster based analyses. fasterraster provides a cython wrapper for optimized openMP c code. Data objects are handled by numpy allowing for straightforward memory management. Currently only computation of visual/morphological features have been implemented however this is open to expansion in the future. All code is still in development and thus it is recommended to test fully before use.

Installation

fasterraster has currently been tested on Linux and Microsoft windows operating systems. You will need python>=3.6 installed. It is recommended to install fasterraster within a virtual environment.

Install using pip

To install fasterraster from PyPI using pip:

pip install fasterraster

Install from source

To build fasterraster from source, download this repository and run:

python3 setup.py build_ext --inplace

Note: You will need to have the required build dependencies installed.

Example

import timeit
import numpy as np
import fasterraster as fr
from pathlib import Path

NTESTS = 10

# Load a .bil file containing a DEM
fname = Path('./test_data/dem.bil')
dem = fr.read(fname)

# regular python implementation of hillshade function
# from https://www.neonscience.org/resources/learning-hub/tutorials/create-hillshade-py
def py_hillshade(dem, cell_size, azimuth=330, altitude=30):
    azimuth = 360.0 - azimuth

    dem = dem / cell_size
    x, y = np.gradient(dem)
    slope = np.pi/2. - np.arctan(np.sqrt(x*x + y*y))
    aspect = np.arctan2(-x, y)
    azimuthrad = azimuth*np.pi/180.
    altituderad = altitude*np.pi/180.
 
    shaded = np.sin(altituderad)*np.sin(slope)
    + np.cos(altituderad)*np.cos(slope) * np.cos(
    (azimuthrad - np.pi/2.) - aspect)
    
    return 255*(shaded + 1)/2

# Time hillshade computation using regular python
time = timeit.timeit(lambda: py_hillshade(dem.raster, dem.XDIM), number=NTESTS)
print(f'python hillshade averaged {time/NTESTS:.3f} seconds')

# Time hillshade computation using fasterraster for num-threads
num_threads = [1,2,4,8]
for numt in num_threads:
    time = timeit.timeit(lambda: fr.hillshade_faster_mp(dem.raster, numt), number=NTESTS)
    print(f'hillshade averaged {time/NTESTS:.3f} seconds for {numt} threads')

Example output:

python hillshade averaged 2.880 seconds
hillshade averaged 0.081 seconds for 1 threads
hillshade averaged 0.041 seconds for 2 threads
hillshade averaged 0.034 seconds for 4 threads
hillshade averaged 0.024 seconds for 8 threads