RRaster

Simple and readable raster manipulation with Python.

This package is an attempt to create a set of user-friendly functions for reprojecting and manipulating rasters with Python. The syntax is meant to mirror that of the R Raster library.

Dependencies

• Python 3.8+
• numpy 1.3+
• rasterio 1.2.6+
• matplotlib 3.4.1+

Installation

It is recommended to use this package within a virtualenv.

$ python3 -m venv venv
$ source venv/bin/activate

# The RRaster package can be installed with pip
$ pip install rraster

# Alternatively, the development version of the package can be installed from 
$ pip install -r requirements.txt
$ pip install .

Examples

The current operations supported by RRaster are reprojection, reduction, writing to disk, and basic raster calculations (addition, subtraction, etc.). Below, a gridMet precipitation raster is projected to the EASE-2 grid and some basic manipulations are done. Although this syntax may be less Pythonic, I find it much easier to read and remember than the typical rasterio syntax.

import numpy as np
from pathlib import Path
import pkg_resources
from rraster.Raster import Raster, RasterStack

# Find path to data within the package
pth = Path(pkg_resources.resource_filename('rraster', 'data'))

# Read in rasters
r1 = Raster(pth / '19990601_pr.tif')
r1 = Raster(pth / 'template.tif')

# Reproject with one line of code
reproj = r1.reproject(r2, method='bilinear')

# Do some stuff to the new raster
reproj = reproj * r2 # Possible because the two rasters now share the same projection and resolution.
reproj /= 18

# Make a RasterStack 
stk = RasterStack(reproj, r2)
# Calculate the pixel-wise mean
new_raster = stk.reduce(np.mean, axis=0)

# Save the new raster to disk.
new_raster.write('./example.tif')