prepared step2.R for first tests (old code, reindented, all in one file)

[epclust.git] / epclust / R / stage2.R

#point avec Jairo:
#rentrer dans code C cwt continue Rwave
#passer partie sowas à C
#fct qui pour deux series (ID, medoides) renvoie distance WER (Rwave ou à moi)
#transformee croisee , smoothing lissage 3 composantes , + calcul pour WER
#attention : code fait pour des series temps desynchronisees ! (deltat, dt == 1,2 ...)
#determiner nvoice noctave (entre octave + petit et + grand)

library("Rwave")

#Entrée : courbes synchrones, soit après étape 1 itérée, soit après chaqure étape 1
#TODO: bout de code qui calcule les courbes synchrones après étapes 1+2 à partir des ID médoïdes

#toCWT: (aux)
##NOTE: renvoie une matrice 3D
toCWT  <- function(X, sw=  0,  tw=  0, swabs= 0, nvoice= 12, noctave= 5, s0= 2, w0= 2*pi,
	lt= 24, dt= 0.5, spectra = FALSE, smooth = TRUE, scaled  = FALSE, scalevector)
{
	noctave  <- adjust.noctave(lt, dt, s0, tw, noctave)
	if(missing(scalevector))
		scalevector  <- 2^(0:(noctave * nvoice) / nvoice) * s0
	res <- lapply(1:nrow(X), function(n) {
		tsX <- ts( X[n,] )
		tsCent <- tsX - mean(tsX)
		if(scaled)
			tsCent <- ts(scale(tsCent))
		tsCent.cwt <- cwt.ts(tsCent, s0, noctave, nvoice, w0)
		tsCent.cwt
	})
	if( spectra )
		res <- lapply(res, function(l) Mod(l)^2 )
	if( smooth  )
		res <- lapply(res, smCWT, swabs = swabs, tw = tw, dt = dt, scalevector = scalevector)
	resArray <- array(NA, c(nrow(res[[1]]), ncol(res[[1]]), length(res)))
	for( l in 1:length(res) )
		resArray[ , , l] <- res[[l]]
	resArray
}

#from sowas
cwt.ts <- function(ts,s0,noctave=5,nvoice=10,w0=2*pi)
{
	if (class(ts)!="ts")
		stop("# This function needs a time series object as input. You may construct this by using the function ts(data,start,deltat). Try '?ts' for help.\n")

	t=time(ts)
	dt=t[2]-t[1]
	s0unit=s0/dt*w0/(2*pi)
	s0log=as.integer((log2(s0unit)-1)*nvoice+1.5)
	if (s0log<1)
	{
		cat(paste("# s0unit = ",s0unit,"\n",sep=""))
		cat(paste("# s0log  = ",s0log,"\n",sep=""))
		cat("# s0 too small for w0! \n")
	}
	totnoct=noctave+as.integer(s0log/nvoice)+1

	#cwt from package Rwave
	totts.cwt=cwt(ts,totnoct,nvoice,w0,plot=0)
	ts.cwt=totts.cwt[,s0log:(s0log+noctave*nvoice)]

	#Normalization
	sqs <- sqrt(2^(0:(noctave*nvoice)/nvoice)*s0)
	smat <- matrix(rep(sqs,length(t)),nrow=length(t),byrow=TRUE)

	ts.cwt*smat
}

#NOTE: vect2mat = as.matrix ?! (dans aux.R)
vect2mat <- function(vect)
{
	vect <- as.vector(vect)
	matrix(vect[-(1:2)], delta, lscvect)
}

#fonction smCWT (dans aux.R)
smCWT <- function(CWT, sw=  0,  tw=  0, swabs= 0, nvoice= 12, noctave= 2, s0= 2, w0= 2*pi,
	lt= 24, dt= 0.5, scalevector )
{
#noctave  <- adjust.noctave(lt, dt, s0, tw, noctave)
#scalevector  <- 2^(0:(noctave * nvoice) / nvoice) * s0
	wsp     <- Mod(CWT)
	smwsp   <- smooth.matrix(wsp, swabs)
	smsmwsp <- smooth.time(smwsp, tw, dt, scalevector)
	smsmwsp
}

#dans sowas.R (...donc on ne lisse pas à ce niveau ?)
smooth.matrix <- function(wt,swabs)
{
	if (swabs != 0)
	{
		smwt <- t(filter(t(wt),rep(1,2*swabs+1)/(2*swabs+1)))
	} else
	{
		smwt <- wt
	}
	smwt
}

smooth.time <- function(wt,tw,dt,scalevector)
{
	smwt <- wt
	if (tw != 0)
	{
		for (i in 1:length(scalevector))
		{
			twi <- as.integer(scalevector[i]*tw/dt)
			smwt[,i] <- filter(wt[,i],rep(1,2*twi+1)/(2*twi+1))
		}
	}
	smwt
}

step2 = function(conso)
{
	#(Benjamin)
	#à partir de là, "conso" == courbes synchrones
	n     <- nrow(conso)
	delta <- ncol(conso)

	#17000 colonnes coeff 1, puis 17000 coeff 2... [non : dans chaque tranche du cube]
	# #NOTE: delta et lscvect pourraient etre gardés à part (communs)

	#TODO: automatic tune of these parameters ? (for other users)
	nvoice   <- 4
	# # noctave4 = 2^13 = 8192 half hours ~ 180 days
	noctave4 <- adjust.noctave(N = delta, dt = 1, s0 = 2, tw = 0, noctave = 13)
	# # 4 here represent 2^5 = 32 half-hours ~ 1 day
	scalevector4  <- 2^(4:(noctave4 * nvoice) / nvoice) * 2
	lscvect4      <- length(scalevector4)
	lscvect <- lscvect4  # i should clean my code: werFam demands a lscvect

	# observations node with CWT
	Xcwt4   <- toCWT(conso, noctave = noctave4, dt = 1, scalevector = scalevector4, lt = delta,
		smooth = FALSE, nvoice = nvoice)

	#matrix:
	############Xcwt2 <- matrix(0.0, nrow= n, ncol= 2 + delta * lscvect)
	Xcwt2 <- matrix(NA_complex_, nrow= n, ncol= 2 + length((c(Xcwt4[,,1]))))

	#NOTE: delta et lscvect pourraient etre gardés à part (communs)
	for(i in 1:n)
		Xcwt2[i,] <- c(delta, lscvect, Xcwt4[,,i] / max(Mod(Xcwt4[,,i])) )
	#rm(conso, Xcwt4); gc()

	## _.b WER^2 distances  ########
	Xwer_dist    <- matrix(0.0, n, n)
	for(i in 1:(n - 1))
	{
		mat1   <- vect2mat(Xcwt2[i,])

	for(j in (i + 1):n)
		{
			mat2 <- vect2mat(Xcwt2[j,])
			num     <- Mod(mat1 * Conj(mat2))
			WX      <- Mod(mat1 * Conj(mat1))
			WY      <- Mod(mat2 * Conj(mat2))
			smsmnum <- smCWT(num, scalevector = scalevector4)
			smsmWX  <- smCWT(WX,  scalevector = scalevector4)
			smsmWY  <- smCWT(WY,  scalevector = scalevector4)
			wer2    <- sum(colSums(smsmnum)^2)  /
			sum( sum(colSums(smsmWX) * colSums(smsmWY)) )
			Xwer_dist[i, j] <- sqrt(delta * lscvect * (1 - wer2))
			Xwer_dist[j, i] <- Xwer_dist[i, j]
		}
	}
	diag(Xwer_dist) <- numeric(n)
	Wwer_dist
}