siteData: An example of site data; this is the portion of the data...

In henkelstone/NPEL.Classification: Classifaction and data-handling routines for NPEL Caribou Project

Description

This data frame contains the site data collected in-house, with the corresponding remote sensed data, both Landsat and DEM and their respective derived categories. See the details section for more information on each included variable.

Usage

siteData$'variables'

Arguments

variables

is the variable to extract from the dataset

Format

A data frame with columns as variables and rows as sites.

Details

In order to build classification models the NPEL collected field data. This vector (point) data was overlaid on the raster data, and the value at each point (site) was extracted. This yielded a dataset of sites with both field characteristics and remote sensed data. This dataset is the basis for the model building in this package. What is included here is a subset both of variables and sites that can be used for learning and testing this package. The sites are only those collected in-house at the NPEL; because our project is done in collaboration with several other agencies we combined their data with ours to produce a larger dataset that spanned more geographic area and more ecosite types. Many of the variable s collected are not relevant to building models (e.g. plot number, date, etc.) and have been pruned from this example. Also several of the remote sensed layers did not prove to be sufficiently linearly independent enough to warrant inclusion in the model; several redundant variables are included here for the purposes of demonstrating dimensional analysis but some have been trimmed in order to keep the example manageable.

• block the block it was sampled under, a component of the sampling regime: c('LA','B1','B2',...,'B6') – large area, blocks 1..6

• landClass the pre-sampling classification of this site, i.e. the stratum it was sampled under

• GPSzone UTM zone of this site; 13 for all our sites

• easting UTM easting of this site

• northing UTM northing of this site

• year the year the field site was visited

• slope field recorded slope of the site, useful to compare with DEM data

• aspect field recorded aspect of the site, useful to compare with DEM data

• ecoType the ecotype the site was classed as

• base.1:7 Landsat base layers

• grnns, wetness, brtns the so-called ‘Tasseled-Cap’ variables; they reflect greenness, wetness, and brightness respectively

• dem the basic elevations from the digital elevation model

• slp slope, as derived from the DEM

• asp aspect, as derived from the DEM; note that different algorithms use different values for NA, a reliable filter is that aspect is NA when slope == 0

• hsd hillshade, a measure of how much light the surface receives – derived from the DEM

Source

• Landsat data: available from the U.S. Geological Survey

• DEM data: available from the U.S. Geological Survey National Elevation Dataset

See Also

See extractPoints for a function to extract raster data for these sites, and generateModels for model building.

Examples

library (NPEL.Classification)
data(siteData)
names(siteData)

siteData$ecoType
siteData[,'ecoType']
siteData[,c('easting','northing')]

modelRun <- generateModels (data = siteData,
                            modelTypes = suppModels,
                            fx = formula('ecoType ~ brtns+grnns+wetns+dem+slp+asp+hsd'),
                            grouping = ecoGroup[['domSpecies','transform']])

## Not run: 
vData <- maptools::readShapePoints ('Input/Plots/Plots')         # Read in from a ESRI shapefile
vData$FEC <- as.factor(substr(as.character(vData$FEC),3,6))      # Extract the FEC number
names (vData) <- c('objID','plotNum','block','type','landClass','ecoRegion','GPSzone','etc..') 
## End(Not run)

henkelstone/NPEL.Classification documentation built on May 17, 2019, 3:42 p.m.