shipunov (version 1.5)

Gradd: Classification grid

Description

Adds to the 2D ordination plot small semi-transparent points which make color classification grid

Usage

Gradd(model2var, data2var, spacing=75, trnsp=0.3, pch=20, cex=0.2,
 palette=NULL, type="ids", User.Predict=function(model2var, X) {}, ...)

Arguments

model2var

Model based on 'data2var' (see below).

data2var

Data with exactly 2 variables.

spacing

Space between points.

trnsp

Transparency.

pch

Type of point.

cex

Scale of points.

palette

Palette to use.

type

Type of the model: "ids", "lda", "neuralnet", "tree", or "user" (see examples).

User.Predict

Function to define in case of 'type="user"'.

...

Additional arguments to plot().

Details

'Gradd()' adds to the 2D ordination plot small semi-transparent points which make color classification grid.

Requires model with 'predict' method to be computed first.

Model should use ids (to make colors) and exactly 2 variables with names same as 'data2var' column names, e.g:

model2var <- somefunction(ids ~ ., data=cbind(ids, data2var))

If type="user", uses predefined 'User.Predict(model2var, X)' function which must return factor ids from testing X data.

Please see examples to understand all of these better.

Note that instead of dots, one can use contours, but they are harder to employ because they need membership values in order to calculate borders (places where memberships are equal).

Examples

# NOT RUN {
## SVM:
library(e1071)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.svm.pca <- svm(Species ~ ., data=cbind(iris[5], iris.p))
plot(iris.p, type="n", main="SVM")
Gradd(iris.svm.pca, iris.p)
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## LDA:
library(MASS)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.lda.pca <- lda(Species ~ . , data=cbind(iris[5], iris.p))
plot(iris.p, type="n", main="LDA")
Gradd(iris.lda.pca, iris.p, type="lda")
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## 'tree::tree':
library(tree)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.tree.pca <- tree(Species ~ . , data=cbind(iris[5], iris.p))
plot(iris.p, type="n", main="tree")
Gradd(iris.tree.pca, iris.p, type="tree")
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## randomForest:
library(randomForest)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.rf.pca <- randomForest(Species ~ ., data=cbind(iris[5], iris.p))
plot(iris.p, type="n", main="randomForest")
Gradd(iris.rf.pca, iris.p)
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## naiveBayes:
library(e1071)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.nb.pca <- naiveBayes(Species ~ ., data=cbind(iris[5], iris.p))
plot(iris.p, type="n", main="naiveBayes")
Gradd(iris.nb.pca, iris.p)
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## neuralnet:
library(neuralnet)
iris.p2 <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.p2 <- cbind(iris.p2, Tobin(iris$Species, convert.names=FALSE))
iris.nn.pca <- neuralnet(setosa + versicolor + virginica ~ PC1 + PC2, data=iris.p2,
 hidden=3, lifesign="full")
plot(iris.p, type="n", main="neuralnet")
Gradd(iris.nn.pca, iris.p, type="neuralnet")
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## rpart + MDS for the base plot:
iris.dist <- dist(iris[, 1:4], method="manhattan")
iris.dist[iris.dist == 0] <- abs(jitter(0))
library(MASS)
iris.m <- isoMDS(iris.dist)$points
colnames(iris.m) <- c("Dim1", "Dim2")
library(rpart)
iris.rpart.mds <- rpart(Species ~ . , data=cbind(iris[5], iris.m))
plot(iris.m, type="n", main="rpart + MDS")
Gradd(iris.rpart.mds, iris.m, type="tree")
text(iris.m, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## QDA:
library(MASS)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.qda.pca <- qda(Species ~ . , data=cbind(iris[5], iris.p))
plot(iris.p, type="n", main="QDA")
Gradd(iris.qda.pca, iris.p, type="lda")
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## AdaBoost:
# }
# NOT RUN {
library(adabag)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.ada.pca <- boosting(Species ~ . , data=cbind(iris[5], iris.p)) # slow!
plot(iris.p, type="n", main="AdaBoost")
Gradd(iris.ada.pca, iris.p, type="lda")
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))
# }
# NOT RUN {
## nnet:
library(nnet)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
iris.nnet.pca <- nnet(Species ~ . , data=cbind(iris[5], iris.p), size=4)
plot(iris.p, type="n", main="nnet")
Gradd(iris.nnet.pca, iris.p, type="tree")
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))

## kNN:
library(class)
iris.p <- prcomp(iris[, 1:4], scale=TRUE)$x[, 1:2]
plot(iris.p, type="n", main="kNN")
Gradd(cbind(iris[5], iris.p), iris.p, type="user",
 User.Predict=function(model2var, X) knn(model2var[, 2:3], X, model2var[, 1], k=5))
text(iris.p, col=as.numeric(iris[, 5]), labels=abbreviate(iris[, 5], 1,
 method="both.sides"))
# }