Morpho_LDA: Morpho_LDA

In micdimu/ecoseeds: Seed ecology - germination and morphometric analysis

Description

The function perform Linear Disciminant Analysis with morphometric data to discriminate taxa (at least 3 groups). The results of LDA are the permormance of discrimination with test-training data, and plotted both with 2 and 3 dimension. The 2 dimension figure is automatically saved in 16:9 at 300 dpi.

Usage

Morpho_LDA(Morpho_df, N.iter=100, split_size=0.80)

Arguments

Morpho_df	Morphometric matrix, row=replicates, col=morphometric variable, frist column=names of taxa (e.g. Genus, species, population...)
N.iter	Number of iterations to assest the LDA performance
split_size	Number from 0 to 1 to split training test data

Value

morpho.lda	output from LDA function
accuracy_train	accurancy discrimination with train data
accuracy_test	accurancy discrimination with test data
confusion_train	confusion matrix with train data
confusion_test	confusion matrix with test data
plot_lda	ggplot output

Author(s)

Michele Di Musciano ([e-mail])

Examples

Morpho_data<-read.csv("your_morpho_file.csv", sep=";", header=T)

Morpho_LDA_output<-Morpho_LDA(Morpho_data, N.iter=100, split_size=0.80)

micdimu/ecoseeds documentation built on Aug. 10, 2020, 10:34 p.m.