percent_cover_parallel: Percentage Cover Calculations

In Tomhigg/fracCover: Generate and Validate Fractional Coverage products using Landsat/MODIS

Description

This script reads training data from the CSV file created using the "create_csv" script. The script then uses the X and Y coordinates from the training data file to select the pixel values (predictor values) for each sample point in the input image. The predictor values and the percent cover data from the training data file (response variable) are combined and used as input to the random forests model. After the model is created percent cover predictions are made on the input image to create an output image with percent cover values ranging from 0 to 1.

Usage

percent_cover_parallel(no_cores, inImage, pointdata, outImage, LS.no.data,
  point_CRS)

Arguments

no_cores	number of cores to implement on
pointdata	Data frame of training data from the rf_csv function, does not need to be projected
LS.no.data	No data value for the Landsat image, normally 0
point_CRS	a CRS object for the training data
LS.stack	A stack of Landsat data, ideally generated by the team LUCC package
outImage.	Name and path for the output classified image, use NA if not needed

Value

a raster image classified into percentage coverage

Tomhigg/fracCover documentation built on May 9, 2019, 5:11 p.m.