FCM.new: Running the improved Fuzzy Cluster Method (FCMm) by...

In bishun945/FCMm: Fuzzy Cluster Method Based on the Optimized m Value (Fuzzifier)

Description

An improved version of Fuzzy Cluster Method (FCM) for water spectra data sets.

Usage

FCM.new(
  FDlist,
  K,
  sort.pos = length(FDlist$wv),
  sort.decreasing = FALSE,
  plot.jitter = TRUE,
  fast.mode = FALSE,
  do.stand = TRUE,
  stand = NULL,
  ...
)

Arguments

FDlist	A list from function FuzzifierDetermination
K	Number, cluster number
sort.pos	The position to sort the cluster number. Default as the end position of the input training matrix. It can be NULL if there is no need to re-sort the cluster index.
sort.decreasing	Default as FALSE which means the sort are assigned from small values to large values.
plot.jitter	Logical, choose to plot jitter plot using ggplot2 functions
fast.mode	Logical, FALSE (default)
do.stand	Whether to use standarized data for FCM. Default as TRUE
stand	Deprecated; Now stand = !do.stand
...	Parameters pass to fcm

Value

A list of FCM:

• FD The return list by function FuzzifierDetermination

• res.FCM The optimized FCM result generated by functions in package ppclust

• K Cluster number

• centroids The list of centroids (both at raw and normalized scale) of the cluster result by aggregating each mean.

• plot.jitter A logical value for the option of doing jitter plot by package ggplot2

• fast.mode A logical value for choosing whether to use fast mode

References

• Bi S, Li Y, Xu J, et al. Optical classification of inland waters based on an improved Fuzzy C-Means method[J]. Optics Express, 2019, 27(24): 34838-34856.

• Dembele D. Multi-objective optimization for clustering 3-way gene expression data[J]. Advances in Data Analysis and Classification, 2008, 2(3): 211-225.

See Also

Other Fuzzy cluster functions: FuzzifierDetermination(), apply_FCM_m(), apply_to_image(), cal_memb(), generate_param(), plot_spec_from_df()

Examples

library(FCMm) 
library(ggplot2) 
library(magrittr)
library(stringr)
data("Nechad2015")
x <- Nechad2015[,3:11]
wv <- gsub("X","",names(x)) %>% as.numeric
w <- sample(1:nrow(x), 100) 
x <- x[w, ]
names(x) <- wv
nb = 4 # Obtained from the vignette "Cluster a new dataset by FCMm"
FD <- FuzzifierDetermination(x, wv, stand=FALSE)
result <- FCM.new(FD, nb, fast.mode = TRUE)

bishun945/FCMm documentation built on Oct. 15, 2021, 6:43 p.m.