RFclass_LLOCV: Random Forest Classification with Leave Location Out...

Description Usage Arguments Details Value Author(s) Examples

View source: R/RFclass_LLOCV.R

Description

RF Classification with LLOCV

Usage

1
2
3
4
5
6
7
8
9
RFclass_LLOCV(
  tDat,
  predCol = "default",
  predStk = NULL,
  classCol = "class",
  classLocCol = "class_location",
  nk = NULL,
  Cores = 1
)

Arguments

tDat

data.frame - with values of the predictors (see details)

predCol

numeric - seq of columns with predictor values. By default uses 1:(length(tDat)-1) for tDat format computed by IKARUS::exrct_Traindat

predStk
  • RasterStack - with the predictors.

classCol

character - name of the column containing the class information

classLocCol

character - name of the column containing the class and location information

nk
  • numeric - number for k in spacefolds

Cores

numeric - amount of Cores to exclude from calculation, default = 1

Details

Value

returns a list with the model and the prediction

Author(s)

Andreas Schönberg

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
# load data
require(caret)
require(CAST)
require(doParallel)
require(raster)
require(IKARUS)
lau_Stk <- raster::stack(system.file("extdata","lau_RGB.grd",package = "IKARUS"))
lau_tP <-rgdal::readOGR(system.file("extdata","lau_TrainPolygon.shp",package = "IKARUS"))
# handle CRS string
crs(lau_tP) <- crs(lau_Stk)
### extract values using 'exrct_Tdat' to generate training dataset
tDat <- exrct_Traindat_LLOCV(lau_tP,lau_Stk,classCol="class",locname="location")
# check for class column and predictor columns in input training dataset
head(tDat)
# classification
model1 <- RFclass_LLOCV(tDat = tDat,predCol = "default",predStk = lau_Stk,classCol = "class",nk=5)
# check model
model1$model_LLOCV
# plot prediction
plot(model1$prediction)

SchoenbergA/IKARUS documentation built on Sept. 8, 2021, 11:11 a.m.