diff --git a/R/PM10.R b/R/PM10.R
index b229a5e..fe84dfa 100644
--- a/R/PM10.R
+++ b/R/PM10.R
@@ -1,23 +1,65 @@
-##' TITLE
+##' Real-wold data on PM10 pollution in Rouen area, France
 ##'
-##' DESCRIPTION
-##' 
-##' @format The format is a list of 2 component:
-##' 
-##' $x: A data-frame containing input variables: with 100 obs. of 200
-##' variables ;
-##' 
-##' $y: Output variable: a factor with 2 levels "-1" and "1".
-##' 
-##' @examples
-##' 
-##' ##'
-##' \dontrun{
+##' These data are TEOM (Tapered Element Oscillating Microbalance) PM10
+##' concentrations from 2004 to 2006 (1096 days) measured by Air Normand, and
+##' the associated weather data provided by Météo France, the French national
+##' meteorological service, using six different monitoring sites.
+##'
+##' Six different monitoring stations of the Rouen (Haute Normandie, France)
+##' area are considered. The urban station \code{jus}, the traffic station
+##' \code{gui}, the second most polluted in the region, and \code{gcm} which is
+##' located in an industrial area. In Le Havre, are considered the stations
+##' \code{rep} (the most polluted in the region) and \code{hri} located at the
+##' seaside. Lastly, the station \code{ail} near Dieppe, because it is rural and
+##' coastal, and a priori hardly influenced by social and industrial activity.
+##' Grouping by categories: \code{jus} and \code{hri} are background urban
+##' monitoring sites, \code{gui} and \code{rep} are urban sites close to
+##' traffic, \code{gcm} is industrial and \code{ail} is rural.
+##'
+##' @format Each object is a data frame.
+##'
+##'   The description of the 18 variables is the following (note that for
+##'   \code{gcm} station, only the pollutant SO2 is available, and there is no
+##'   pollutant for \code{ail} station):
 ##' 
+##' \describe{
+##'   \item{PM10}{Daily mean concentration of PM10, in \eqn{\mu}g/m3}
+##'   \item{NO, NO2, SO2}{Daily mean concentration of NO, NO2 , SO2, in 
+##'   \eqn{\mu}g/m3}
+##'   \item{T.min, T.max, T.moy}{Daily minimum, maximum and mean temperature, in
+##'   °C}
+##'   \item{DV.maxvv, DV.dom}{Daily maximum speed and dominant wind direction,
+##'   in ° (for wind direction, 0° corresponds to north)}
+##'   \item{VV.max, VV.moy}{Daily maximum and mean wind speed, in m/s}
+##'   \item{PL.som}{Daily rainfall, in mm}
+##'   \item{HR.min, HR.max, HR.moy}{Daily minimum, maximum and mean relative
+##'   humidity, in \%}
+##'   \item{PA.moy}{Daily mean air pressure, in hPa}
+##'   \item{GTrouen, GTlehavre}{Daily temperature gradient, in °C}
 ##' }
 ##' 
-##' @source   Weston, J., Elisseff, A., Schoelkopf, B., Tipping, M. (2003),
-##' \emph{Use of the zero norm with linear models and Kernel methods},
-##' J. Machine Learn. Res. 3, 1439-1461
-##'
-"jus"
\ No newline at end of file
+##' @source F.-X. Jollois, J.-M. Poggi, B. Portier, \emph{Three non-linear
+##'   statistical methods to analyze PM10 pollution in Rouen area}. CSBIGS 3(1):
+##'   1-17, 2009
+##' 
+##' @docType data
+##' @name PM10
+NULL
+
+##' @rdname PM10
+"ail"
+
+##' @rdname PM10
+"gcm"
+
+##' @rdname PM10
+"gui"
+
+##' @rdname PM10
+"hri"
+
+##' @rdname PM10
+"jus"
+
+##' @rdname PM10
+"rep"
\ No newline at end of file
diff --git a/R/toys.R b/R/toys.R
index bd2a133..9797455 100644
--- a/R/toys.R
+++ b/R/toys.R
@@ -3,28 +3,31 @@
 ##' \code{toys} is a simple simulated dataset of a binary classification
 ##' problem, introduced by Weston et.al..
 ##'
-##' It is an equiprobable two class problem, Y belongs to {-1,1}, with six
+##' It is an equiprobable two class problem, Y belongs to \{-1,1\}, with six
 ##' true  variables, the others being some noise.
 ##' The simulation model is defined through the conditional distribution
-##' of the Xi for Y=y:
+##' of the X_i for Y=y:
 ##' 
-##' with probability 0.7, X^j ~ N(yj,1) for j=1,2,3 and
+##' \itemize{
+##'  \item with probability 0.7, X^j ~ N(yj,1) for j=1,2,3 and
 ##' X^j ~ N(0,1) for j=4,5,6 ;
 ##' 
-##' with probability 0.3, X^j ~ N(0,1) for j=1,2,3 and
+##'  \item with probability 0.3, X^j ~ N(0,1) for j=1,2,3 and
 ##' X^j ~ N(y(j-3),1) for j=4,5,6 ;
 ##' 
-##' the other variables are noise, X^j ~ N(0,1)
+##'  \item the other variables are noise, X^j ~ N(0,1)
 ##' for j=7,\dots,p.
+##' }
 ##' 
 ##' After simulation, the obtained variables are finally standardized.
 ##' 
-##' @format The format is a list of 2 component:
-##' 
-##' $x: A data-frame containing input variables: with 100 obs. of 200
-##' variables ;
+##' @format The format is a list of 2 components:
 ##' 
-##' $y: Output variable: a factor with 2 levels "-1" and "1".
+##' \describe{
+##'  \item{x}{a dataframe containing input variables: with 100 obs. of 200
+##' variables}
+##'  \item{y}{output variable: a factor with 2 levels "-1" and "1"}
+##'  }
 ##' 
 ##' @examples
 ##' data(toys)
diff --git a/data/ail_comp.rda b/data/ail_comp.rda
deleted file mode 100644
index 93c03a6..0000000
Binary files a/data/ail_comp.rda and /dev/null differ
diff --git a/data/gcm_comp.rda b/data/gcm_comp.rda
deleted file mode 100644
index 4608afe..0000000
Binary files a/data/gcm_comp.rda and /dev/null differ
diff --git a/data/gui_comp.rda b/data/gui_comp.rda
deleted file mode 100644
index babafae..0000000
Binary files a/data/gui_comp.rda and /dev/null differ
diff --git a/data/hri_comp.rda b/data/hri_comp.rda
deleted file mode 100644
index e35ad3a..0000000
Binary files a/data/hri_comp.rda and /dev/null differ
diff --git a/data/jus_comp.rda b/data/jus_comp.rda
deleted file mode 100644
index 4b41219..0000000
Binary files a/data/jus_comp.rda and /dev/null differ
diff --git a/data/rep_comp.rda b/data/rep_comp.rda
deleted file mode 100644
index bf77cd1..0000000
Binary files a/data/rep_comp.rda and /dev/null differ
diff --git a/man/PM10.Rd b/man/PM10.Rd
new file mode 100644
index 0000000..1bfc9db
--- /dev/null
+++ b/man/PM10.Rd
@@ -0,0 +1,70 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/PM10.R
+\docType{data}
+\name{PM10}
+\alias{PM10}
+\alias{ail}
+\alias{gcm}
+\alias{gui}
+\alias{hri}
+\alias{jus}
+\alias{rep}
+\title{Real-wold data on PM10 pollution in Rouen area, France}
+\format{Each object is a data frame.
+
+  The description of the 18 variables is the following (note that for
+  \code{gcm} station, only the pollutant SO2 is available, and there is no
+  pollutant for \code{ail} station):
+
+\describe{
+  \item{PM10}{Daily mean concentration of PM10, in \eqn{\mu}g/m3}
+  \item{NO, NO2, SO2}{Daily mean concentration of NO, NO2 , SO2, in 
+  \eqn{\mu}g/m3}
+  \item{T.min, T.max, T.moy}{Daily minimum, maximum and mean temperature, in
+  °C}
+  \item{DV.maxvv, DV.dom}{Daily maximum speed and dominant wind direction,
+  in ° (for wind direction, 0° corresponds to north)}
+  \item{VV.max, VV.moy}{Daily maximum and mean wind speed, in m/s}
+  \item{PL.som}{Daily rainfall, in mm}
+  \item{HR.min, HR.max, HR.moy}{Daily minimum, maximum and mean relative
+  humidity, in \%}
+  \item{PA.moy}{Daily mean air pressure, in hPa}
+  \item{GTrouen, GTlehavre}{Daily temperature gradient, in °C}
+}}
+\source{
+F.-X. Jollois, J.-M. Poggi, B. Portier, \emph{Three non-linear
+  statistical methods to analyze PM10 pollution in Rouen area}. CSBIGS 3(1):
+  1-17, 2009
+}
+\usage{
+ail
+
+gcm
+
+gui
+
+hri
+
+jus
+
+rep
+}
+\description{
+These data are TEOM (Tapered Element Oscillating Microbalance) PM10
+concentrations from 2004 to 2006 (1096 days) measured by Air Normand, and
+the associated weather data provided by Météo France, the French national
+meteorological service, using six different monitoring sites.
+}
+\details{
+Six different monitoring stations of the Rouen (Haute Normandie, France)
+area are considered. The urban station \code{jus}, the traffic station
+\code{gui}, the second most polluted in the region, and \code{gcm} which is
+located in an industrial area. In Le Havre, are considered the stations
+\code{rep} (the most polluted in the region) and \code{hri} located at the
+seaside. Lastly, the station \code{ail} near Dieppe, because it is rural and
+coastal, and a priori hardly influenced by social and industrial activity.
+Grouping by categories: \code{jus} and \code{hri} are background urban
+monitoring sites, \code{gui} and \code{rep} are urban sites close to
+traffic, \code{gcm} is industrial and \code{ail} is rural.
+}
+\keyword{datasets}
diff --git a/man/jus.Rd b/man/jus.Rd
deleted file mode 100644
index db32594..0000000
--- a/man/jus.Rd
+++ /dev/null
@@ -1,32 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/PM10.R
-\docType{data}
-\name{jus}
-\alias{jus}
-\title{TITLE}
-\format{The format is a list of 2 component:
-
-$x: A data-frame containing input variables: with 100 obs. of 200
-variables ;
-
-$y: Output variable: a factor with 2 levels "-1" and "1".}
-\source{
-Weston, J., Elisseff, A., Schoelkopf, B., Tipping, M. (2003),
-\emph{Use of the zero norm with linear models and Kernel methods},
-J. Machine Learn. Res. 3, 1439-1461
-}
-\usage{
-jus
-}
-\description{
-DESCRIPTION
-}
-\examples{
-
-##'
-\dontrun{
-
-}
-
-}
-\keyword{datasets}
diff --git a/man/toys.Rd b/man/toys.Rd
index 63b7d29..d3d066e 100644
--- a/man/toys.Rd
+++ b/man/toys.Rd
@@ -4,12 +4,13 @@
 \name{toys}
 \alias{toys}
 \title{A simulated dataset called toys data}
-\format{The format is a list of 2 component:
+\format{The format is a list of 2 components:
 
-$x: A data-frame containing input variables: with 100 obs. of 200
-variables ;
-
-$y: Output variable: a factor with 2 levels "-1" and "1".}
+\describe{
+ \item{x}{a dataframe containing input variables: with 100 obs. of 200
+variables}
+ \item{y}{output variable: a factor with 2 levels "-1" and "1"}
+ }}
 \source{
 Weston, J., Elisseff, A., Schoelkopf, B., Tipping, M. (2003),
 \emph{Use of the zero norm with linear models and Kernel methods},
@@ -20,19 +21,21 @@ J. Machine Learn. Res. 3, 1439-1461
 problem, introduced by Weston et.al..
 }
 \details{
-It is an equiprobable two class problem, Y belongs to {-1,1}, with six
+It is an equiprobable two class problem, Y belongs to \{-1,1\}, with six
 true  variables, the others being some noise.
 The simulation model is defined through the conditional distribution
-of the Xi for Y=y:
+of the X_i for Y=y:
 
-with probability 0.7, X^j ~ N(yj,1) for j=1,2,3 and
+\itemize{
+ \item with probability 0.7, X^j ~ N(yj,1) for j=1,2,3 and
 X^j ~ N(0,1) for j=4,5,6 ;
 
-with probability 0.3, X^j ~ N(0,1) for j=1,2,3 and
+\item with probability 0.3, X^j ~ N(0,1) for j=1,2,3 and
 X^j ~ N(y(j-3),1) for j=4,5,6 ;
 
-the other variables are noise, X^j ~ N(0,1)
+\item the other variables are noise, X^j ~ N(0,1)
 for j=7,\dots,p.
+}
 
 After simulation, the obtained variables are finally standardized.
 }