diff --git a/R/kfino.R b/R/kfino.R
index 1a2db71d8b0feff408b82d07f8744d9923177dfd..b02e62c2a56add68de6dc3d92bf5479dd0a8f9ab 100644
--- a/R/kfino.R
+++ b/R/kfino.R
@@ -185,10 +185,10 @@ kfino_fit<-function(datain,Tvar,Yvar,
#WARNING WARNING: AU lieu de calculer L qui est arrondi à 0,
# je calcule 10^N fois Lu et 10^N q. Au total pr chaque donnée
- # je multiplie par 100 mais comme c'est l'ordre de grandeur de loioutlier()
+ # je multiplie par 100 mais comme c'est l'ordre de grandeur de doutlier()
# ca ne me parait pas disproportionné.
# Au lieu de 10 et 10 je fais simplement sqrt(expertMax - expertMin)
- dix=sqrt(expertMax - expertMin)
+ scalingC=sqrt(expertMax - expertMin)
#------------------------------------------------------------------------
# Optimisation on Initial parameters or not
@@ -227,7 +227,7 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_m0=sigma2_m0,
sigma2_pp=sigma2_pp,
K=K),
- Y=Y,Tps=Tps,N=N,dix=dix,kappaOpt=kappaOpt)
+ Y=Y,Tps=Tps,N=N,scalingC=scalingC,kappaOpt=kappaOpt)
for (m0 in seq(bornem0[1],bornem0[2],2) ){
for (mm in seq((m0-5),(m0+20),2) ){
@@ -248,7 +248,7 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_m0=sigma2_m0,
sigma2_pp=sigma2_pp,
K=K),
- Y=Y,Tps=Tps,N=N,dix=dix,kappaOpt=kappaOpt)
+ Y=Y,Tps=Tps,N=N,scalingC=scalingC,kappaOpt=kappaOpt)
if (V > Vopt){
Vopt=V
m0opt=m0
@@ -311,7 +311,7 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_mm=sigma2_mm,
sigma2_pp=sigma2_pp,
K=K),
- kappaOpt=kappaOpt, Y=Y,Tps=Tps,N=N,dix=dix)
+ kappaOpt=kappaOpt, Y=Y,Tps=Tps,N=N,scalingC=scalingC)
diff_m0=abs(m0_tmp - Res_EM$m0[[1]])
diff_p=abs(p_tmp - Res_EM$pp)
diff_mm=abs(m_tmp - Res_EM$mm[[1]])
@@ -345,7 +345,7 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_mm=sigma2_mm,
sigma2_pp=sigma2_pp,
K=K),
- kappaOpt=kappaOpt, Y=Y,Tps=Tps,N=N,dix=dix)
+ kappaOpt=kappaOpt, Y=Y,Tps=Tps,N=N,scalingC=scalingC)
diff_m0=abs(m0_tmp - Res_EM$m0[[1]])
diff_p=abs(p_tmp - Res_EM$pp)
diff_mm=abs(m_tmp - Res_EM$mm[[1]])
@@ -384,7 +384,7 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_mm=sigma2_mm,
sigma2_pp=sigma2_pp,
K=K),
- kappaOpt=kappaOpt, Y=Y,Tps=Tps,N=N,dix=dix)
+ kappaOpt=kappaOpt, Y=Y,Tps=Tps,N=N,scalingC=scalingC)
diff_m0=abs(m0_tmp - Res_EM$m0[[1]])
diff_p=abs(p_tmp - Res_EM$pp)
diff_mm=abs(m_tmp - Res_EM$mm[[1]])
@@ -430,7 +430,8 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_m0=sigma2_m0,
sigma2_pp=sigma2_pp,
K=K),
- Y=Y,Tps=Tps,N=N,dix=dix,kappaOpt=kappaOpt)$likelihood
+ Y=Y,Tps=Tps,N=N,scalingC=scalingC,
+ kappaOpt=kappaOpt)$likelihood
if (Vopt_low > Vopt){
m0opt<-m0opt_low
@@ -479,7 +480,7 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_m0=sigma2_m0,
sigma2_pp=sigma2_pp,
K=K),
- Y=Y,Tps=Tps,N=N,dix=dix,kappaOpt=kappaOpt)
+ Y=Y,Tps=Tps,N=N,scalingC=scalingC,kappaOpt=kappaOpt)
for (m0 in seq(bornem0[1],bornem0[2],2) ){
for (mm in seq((m0-5),(m0+20),2) ){
for (p in seqp){
@@ -493,7 +494,7 @@ kfino_fit<-function(datain,Tvar,Yvar,
sigma2_m0=sigma2_m0,
sigma2_pp=sigma2_pp,
K=K),
- Y=Y,Tps=Tps,N=N,dix=dix,kappaOpt=kappaOpt)
+ Y=Y,Tps=Tps,N=N,scalingC=scalingC,kappaOpt=kappaOpt)
if (V > Vopt){
Vopt=V
diff --git a/R/utils_functions.R b/R/utils_functions.R
index ef18f9cf312e67575796599618504778f112d71a..0d94b02e21964623129cfcd50f1ced5299d5dc84 100644
--- a/R/utils_functions.R
+++ b/R/utils_functions.R
@@ -33,15 +33,16 @@ doutlier<-function(y,
#--------------------------------------------------------------------------
#' KBO_known a function to calculate a likelihood on given parameters
#'
-#' @param param list, a list of 10 input parameters for mm, pp and m0
+#' @param param list, see initial parameter list in \code{kfino_fit}
#' @param threshold numeric, threshold for CI, default 0.5
-#' @param kappa numeric, truncation setting, default 10
+#' @param kappa numeric, truncation setting for likelihood optimization,
+#' default 10
#' @param Y character, name of the numeric variable to predict in the
#' data.frame datain
#' @param Tps character, time column name in the data.frame datain, a
#' numeric vector.
#' Tvar can be expressed as a proportion of day in seconds
-#' @param N numeric, length of Y
+#' @param N numeric, length of the numeric vector of Y values
#'
#' @details uses the same input parameter list than the main function
#' @return a list
@@ -229,15 +230,16 @@ KBO_known<-function(param,threshold,kappa=10,Y,Tps,N){
#' KBO_L a function to calculate a likelihood on initial parameters
#' optimized by a grid search
#'
-#' @param param a list of 10 input parameters mm, pp and m0
-#' @param kappaOpt numeric, truncation setting, default 7
+#' @param param list, see initial parameter list in \code{kfino_fit}
+#' @param kappaOpt numeric, truncation setting for initial parameters'
+#' optimization, default 7
#' @param Y character, name of the numeric variable to predict in the
#' data.frame datain
#' @param Tps character, time column name in the data.frame datain, a
#' numeric vector.
#' Tvar can be expressed as a proportion of day in seconds
-#' @param N numeric, length of Y
-#' @param dix num
+#' @param N numeric, length of the numeric vector of Y values
+#' @param scalingC numeric, scaling constant
#'
#' @details uses the same input parameter list than the main function
#' @return a likelihood
@@ -258,10 +260,10 @@ KBO_known<-function(param,threshold,kappa=10,Y,Tps,N){
#' sigma2_mm=0.05,
#' sigma2_pp=5,
#' K=2,
-#' seqp=seq(0.5,0.7,0.1))
+#' seqp=seq(0.5,0.7,0.1))
#' print(Y)
-#' KBO_L(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,dix=6)
-KBO_L<-function(param,kappaOpt=7,Y,Tps,N,dix){
+#' KBO_L(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,scalingC=6)
+KBO_L<-function(param,kappaOpt=7,Y,Tps,N,scalingC){
# load objects
mm=param[["mm"]]
pp=param[["pp"]]
@@ -293,8 +295,8 @@ KBO_L<-function(param,kappaOpt=7,Y,Tps,N,dix){
m=c(m0,m1)
p=c(p0,p1)
sigma2=c(sigma2_m0,sigma1)
- L=dix*c(l0,loinorm1) #Attention *2 pr le grandir
- q=c(1,1) #attention
+ L=scalingC*c(l0,loinorm1) # increase it with the scaling constant
+ q=c(1,1)
#iteration (1.1.2)
#-----------------------
@@ -327,8 +329,8 @@ KBO_L<-function(param,kappaOpt=7,Y,Tps,N,dix){
sigma2=sigma2new
p=pnew/sum(pnew)
- L=dix*Lnew
- q=dix*qnew
+ L=scalingC*Lnew
+ q=scalingC*qnew
}
# after truncation
@@ -364,8 +366,8 @@ KBO_L<-function(param,kappaOpt=7,Y,Tps,N,dix){
m=mnew[selection]
sigma2=sigma2new[selection]
p=pnew[selection]/sum(pnew[selection])
- L=dix*Lnew[selection]
- q=dix*qnew[selection]
+ L=scalingC*Lnew[selection]
+ q=scalingC*qnew[selection]
}
Vraisemblance=L%*%q
@@ -377,15 +379,16 @@ KBO_L<-function(param,kappaOpt=7,Y,Tps,N,dix){
#' KBO_EM a function to calculate a likelihood on initial parameters
#' optimized by an Expectation-Maximization (EM) algorithm
#'
-#' @param param a list of 10 input parameters mm, pp and m0
-#' @param kappaOpt numeric, truncation setting, default 7
+#' @param param list, see initial parameter list in \code{kfino_fit}
+#' @param kappaOpt numeric, truncation setting for initial parameters'
+#' optimization, default 7
#' @param Y character, name of the numeric variable to predict in the
#' data.frame datain
#' @param Tps character, time column name in the data.frame datain, a
#' numeric vector.
#' Tvar can be expressed as a proportion of day in seconds
-#' @param N numeric, length of Y
-#' @param dix num
+#' @param N numeric, length of the numeric vector of Y values
+#' @param scalingC numeric, scaling constant
#'
#' @details uses the same input parameter list than the main function
#' @return a list:
@@ -412,10 +415,10 @@ KBO_L<-function(param,kappaOpt=7,Y,Tps,N,dix){
#' sigma2_mm=0.05,
#' sigma2_pp=5,
#' K=2,
-#' seqp=seq(0.5,0.7,0.1))
+#' seqp=seq(0.5,0.7,0.1))
#' print(Y)
-#' KBO_EM(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,dix=6)
-KBO_EM<-function(param,kappaOpt,Y,Tps,N,dix){
+#' KBO_EM(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,scalingC=6)
+KBO_EM<-function(param,kappaOpt,Y,Tps,N,scalingC){
# load objects
mm<-param[["mm"]]
pp<-param[["pp"]]
@@ -447,8 +450,8 @@ KBO_EM<-function(param,kappaOpt,Y,Tps,N,dix){
m=c(m0,m1)
p=c(p0,p1)
sigma2=c(sigma2_m0,sigma1)
- L=dix*c(l0,loinorm1) #Attention *2 pr le grandir
- q=c(1,1) #attention
+ L=scalingC*c(l0,loinorm1)
+ q=c(1,1)
a=c(1,sigma2_pp/(sigma2_m0+sigma2_pp))
b=c(0,0)
@@ -521,8 +524,8 @@ KBO_EM<-function(param,kappaOpt,Y,Tps,N,dix){
Znew[(2^k+1):(2^(k+1)),]=cbind(Z, rep(1,2^k))
Z=Znew
- L=dix*Lnew # fois 2 pr le grandir
- q=dix*qnew
+ L=scalingC*Lnew # fois 2 pr le grandir
+ q=scalingC*qnew
}
# after truncation
@@ -573,8 +576,8 @@ KBO_EM<-function(param,kappaOpt,Y,Tps,N,dix){
m=mnew[selection]
sigma2=sigma2new[selection]
p=pnew[selection]/sum(pnew[selection])
- L=dix*Lnew[selection] #fois 2 pr le grandir
- q=dix*qnew[selection]
+ L=scalingC*Lnew[selection]
+ q=scalingC*qnew[selection]
Znew=matrix(rep(0,(k+1)*2^(kappa+1)), ncol=k+1, nrow=2^(kappa+1), byrow=TRUE)
Znew[1:2^kappa,]=cbind(Z, rep(0,2^kappa))
diff --git a/man/KBO_EM.Rd b/man/KBO_EM.Rd
index 1b5835a667468928b63322c1c1189c66ec6f3fdc..591ebf3e361991e8ffd819048ed6afe6075ee3d6 100644
--- a/man/KBO_EM.Rd
+++ b/man/KBO_EM.Rd
@@ -5,12 +5,13 @@
\title{KBO_EM a function to calculate a likelihood on initial parameters
optimized by an Expectation-Maximization (EM) algorithm}
\usage{
-KBO_EM(param, kappaOpt, Y, Tps, N, dix)
+KBO_EM(param, kappaOpt, Y, Tps, N, scalingC)
}
\arguments{
-\item{param}{a list of 10 input parameters mm, pp and m0}
+\item{param}{list, see initial parameter list in \code{kfino_fit}}
-\item{kappaOpt}{numeric, truncation setting, default 7}
+\item{kappaOpt}{numeric, truncation setting for initial parameters'
+optimization, default 7}
\item{Y}{character, name of the numeric variable to predict in the
data.frame datain}
@@ -19,9 +20,9 @@ data.frame datain}
numeric vector.
Tvar can be expressed as a proportion of day in seconds}
-\item{N}{numeric, length of Y}
+\item{N}{numeric, length of the numeric vector of Y values}
-\item{dix}{num}
+\item{scalingC}{numeric, scaling constant}
}
\value{
a list:
@@ -54,7 +55,7 @@ param2<-list(m0=41,
sigma2_mm=0.05,
sigma2_pp=5,
K=2,
- seqp=seq(0.5,0.7,0.1))
+ seqp=seq(0.5,0.7,0.1))
print(Y)
-KBO_EM(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,dix=6)
+KBO_EM(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,scalingC=6)
}
diff --git a/man/KBO_L.Rd b/man/KBO_L.Rd
index a951cb6f39db6d8628725d2b64b67b391794bb22..f7c4b533c004500267a7ef342365d7a4ceb04e46 100644
--- a/man/KBO_L.Rd
+++ b/man/KBO_L.Rd
@@ -5,12 +5,13 @@
\title{KBO_L a function to calculate a likelihood on initial parameters
optimized by a grid search}
\usage{
-KBO_L(param, kappaOpt = 7, Y, Tps, N, dix)
+KBO_L(param, kappaOpt = 7, Y, Tps, N, scalingC)
}
\arguments{
-\item{param}{a list of 10 input parameters mm, pp and m0}
+\item{param}{list, see initial parameter list in \code{kfino_fit}}
-\item{kappaOpt}{numeric, truncation setting, default 7}
+\item{kappaOpt}{numeric, truncation setting for initial parameters'
+optimization, default 7}
\item{Y}{character, name of the numeric variable to predict in the
data.frame datain}
@@ -19,9 +20,9 @@ data.frame datain}
numeric vector.
Tvar can be expressed as a proportion of day in seconds}
-\item{N}{numeric, length of Y}
+\item{N}{numeric, length of the numeric vector of Y values}
-\item{dix}{num}
+\item{scalingC}{numeric, scaling constant}
}
\value{
a likelihood
@@ -48,7 +49,7 @@ param2<-list(m0=41,
sigma2_mm=0.05,
sigma2_pp=5,
K=2,
- seqp=seq(0.5,0.7,0.1))
+ seqp=seq(0.5,0.7,0.1))
print(Y)
-KBO_L(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,dix=6)
+KBO_L(param=param2,kappaOpt=7,Y=Y,Tps=Tps,N=N,scalingC=6)
}
diff --git a/man/KBO_known.Rd b/man/KBO_known.Rd
index 641f63812cabe3468fa75c218427b1cf22c4507e..e28e51fafc0f7ee221d2d15ee76b7935cb19fa78 100644
--- a/man/KBO_known.Rd
+++ b/man/KBO_known.Rd
@@ -7,11 +7,12 @@
KBO_known(param, threshold, kappa = 10, Y, Tps, N)
}
\arguments{
-\item{param}{list, a list of 10 input parameters for mm, pp and m0}
+\item{param}{list, see initial parameter list in \code{kfino_fit}}
\item{threshold}{numeric, threshold for CI, default 0.5}
-\item{kappa}{numeric, truncation setting, default 10}
+\item{kappa}{numeric, truncation setting for likelihood optimization,
+default 10}
\item{Y}{character, name of the numeric variable to predict in the
data.frame datain}
@@ -20,7 +21,7 @@ data.frame datain}
numeric vector.
Tvar can be expressed as a proportion of day in seconds}
-\item{N}{numeric, length of Y}
+\item{N}{numeric, length of the numeric vector of Y values}
}
\value{
a list