From 61cc1d00d7dfb75bf70aaa9e255210645307e6cf Mon Sep 17 00:00:00 2001
From: Rolf Simoes <rolfsimoes@gmail.com>
Date: Thu, 17 Aug 2017 09:57:05 -0300
Subject: [PATCH 1/7] Implementing ML functions

---
 R/sits_machine_learning.R | 156 ++++++++++++++++++++++++++++++++++----
 R/sits_table.R            |   1 +
 2 files changed, 142 insertions(+), 15 deletions(-)

diff --git a/R/sits_machine_learning.R b/R/sits_machine_learning.R
index 135275345..6bae309e4 100644
--- a/R/sits_machine_learning.R
+++ b/R/sits_machine_learning.R
@@ -26,7 +26,7 @@
 #'              In the context of time series classification, the only relevant types are "C-classification" and "nu-classification".
 #'              Please refer to the documentation in that package for more details.
 #'
-#' @param data.tb     a SITS tibble time series with an alignment column
+#' @param data_tb     a SITS tibble time series with an alignment column
 #' @param kernel      the kernel used in training and predicting (options = linear, polynomial, radial basis, sigmoid)
 #' @param degree      exponential of polynomial type kernel
 #' @param coef0	      parameter needed for kernels of type polynomial and sigmoid (default: 0)
@@ -36,20 +36,20 @@
 #' @return result.svm an svm model fit for the input data
 #' @export
 #'
-sits_train_svm <- function(data.tb = NULL, kernel = "linear",
+sits_train_svm <- function(data_tb = NULL, kernel = "linear",
                            degree = 3, coef0 = 0, cost = 100, tolerance = 0.001, epsilon = 0.1){
 
     # is the input data the result of a TWDTW matching function?
-    ensurer::ensure_that(data.tb, "matches" %in% names (.), err_desc = "sits_train_svm: input data does not contain TWDTW matches")
+    ensurer::ensure_that(data_tb, "matches" %in% names (.), err_desc = "sits_train_svm: input data does not contain TWDTW matches")
 
     # Spread TWDTW matches
-    spread.tb <- .sits_spread_matches(data.tb)
+    spread_tb <- .sits_spread_matches(data_tb)
 
-    categories <- names(spread.tb)[-2:0]
+    categories <- names(spread_tb)[-2:0]
     lognomes <- paste0('log(', categories, ')')
     formula1 <- stats::as.formula(paste("factor(reference) ~ ", paste(lognomes, collapse = " + ")))
 
-    result.svm <- e1071::svm(formula1, data = spread.tb,
+    result.svm <- e1071::svm(formula1, data = spread_tb,
                              type = "C-classification", kernel = kernel,
                              degree = degree, epsilon = epsilon, cost = cost)
 }
@@ -80,7 +80,7 @@ sits_train_svm <- function(data.tb = NULL, kernel = "linear",
 #' @export
 #'
 sits_train <- function(data_tb, training_method = sits_svm(formula = sits_formula_logref(predictors_index = -2:0), kernel = "linear",
-                                                           degree = 3, coef0 = 0, tolerance = 0.001, epsilon = 0.1, ...)){
+                                                           degree = 3, coef0 = 0, tolerance = 0.001, epsilon = 0.1)){
 
     # is the input data the result of a TWDTW matching function?
     ensurer::ensure_that(data_tb, "matches" %in% names (.), err_desc = "sits_train: input data does not contain TWDTW matches")
@@ -89,7 +89,7 @@ sits_train <- function(data_tb, training_method = sits_svm(formula = sits_formul
     ensurer::ensure_that(training_method, class(.) == "function", err_desc = "sits_train: train_method is not a valid function")
 
     # Spread TWDTW matches
-    spread_tb <- .sits_spread_matches(data.tb)
+    spread_tb <- .sits_spread_matches(data_tb)
 
     # compute the training method by the given data
     result <- training_method(spread_tb)
@@ -109,7 +109,7 @@ sits_train <- function(data_tb, training_method = sits_svm(formula = sits_formul
 #'
 #' @param data_tb   a SITS tibble time series with an alignment column. If data_tb is NULL, the function returns
 #' a function to be called further to compute svm training model according to given parameters
-#' @param formula   a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_svm)
+#' @param formula   a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
 #' @param kernel    the kernel used in training and predicting (options = linear, polynomial, radial basis, sigmoid)
 #' @param degree    exponential of polynomial type kernel
 #' @param coef0	    parameter needed for kernels of type polynomial and sigmoid (default: 0)
@@ -144,6 +144,132 @@ sits_svm <- function(data_tb = NULL, formula = sits_formula_logref(), kernel = "
     return(result)
 }
 
+#' @title Train SITS classifiction models
+#' @name sits_lda
+#'
+#' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Victor Maus, \email{vwmaus1@@gmail.com}
+#' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Performs a linear discriminant analysis (lda) to classify data.
+#' This function is a front-end to the "lda" method in the "MASS" package.
+#' Please refer to the documentation in that package for more details.
+#'
+#' @param data_tb   a SITS tibble time series with an `matches` column. If `data_tb` is NULL, the function returns
+#' a pre-parameterized function to be called further in order to compute `lda` training model
+#' @param formula   a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
+#' @param ...       other parameters to be passed to MASS::lda function
+#' @return result   either an MASS::lda class or an function prepared that can be called further to compute lda training model
+#' @export
+#'
+sits_lda <- function(data_tb = NULL, formula = sits_formula_logref(), ...) {
+
+    # function that returns MASS::lda model based on a sits sample tibble
+    result_fun <- function(tb){
+
+        # if parameter formula is a function call it passing as argument the input data sample. The function must return a valid formula.
+        if (class(formula) == "function")
+            formula <- formula(tb)
+
+        # call MASS::lda method and return the trained lda model
+        result_lda <- MASS::lda(formula = formula, data = tb, ...)
+        return(result_lda)
+    }
+
+    # if no data is given, we prepare a function to compute lda as model
+    if (is.null(data_tb))
+        result <- result_fun
+    # ...otherwise compute lda model and return it
+    else
+        result <- result_fun(data_tb)
+    return(result)
+}
+
+#' @title Train SITS classifiction models
+#' @name sits_mlr
+#'
+#' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Victor Maus, \email{vwmaus1@@gmail.com}
+#' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Use multinomial log-linear (mlr) fitting model via neural networks to classify data.
+#' This function is a front-end to the "multinom" method in the "nnet" package.
+#' Please refer to the documentation in that package for more details.
+#'
+#' @param data_tb   a SITS tibble time series with an `matches` column. If `data_tb` is NULL, the function returns
+#' a pre-parameterized function to be called further in order to compute `multinom` training model
+#' @param formula   a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
+#' @param ...       other parameters to be passed to nnet::multinom function
+#' @return result   either an nnet::multinom class or an function prepared that can be called further to compute multinom training model
+#' @export
+#'
+sits_mlr <- function(data_tb = NULL, formula = sits_formula_logref(), ...) {
+
+    # function that returns nnet::multinom model based on a sits sample tibble
+    result_fun <- function(tb){
+
+        # if parameter formula is a function call it passing as argument the input data sample. The function must return a valid formula.
+        if (class(formula) == "function")
+            formula <- formula(tb)
+
+        # call nnet::multinom method and return the trained multinom model
+        result_mlr <- nnet::multinom(formula = formula, data = tb, ...)
+        return(result_mlr)
+    }
+
+    # if no data is given, we prepare a function to compute multinom as model
+    if (is.null(data_tb))
+        result <- result_fun
+    # ...otherwise compute multinom model and return it
+    else
+        result <- result_fun(data_tb)
+    return(result)
+}
+
+#' @title Train SITS classifiction models
+#' @name sits_glm
+#'
+#' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Victor Maus, \email{vwmaus1@@gmail.com}
+#' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Use generalized liner model (glm) via penalized maximim likelihood to classify data.
+#' This function is a front-end to the "glmnet" method in the "glmnet" package.
+#' Please refer to the documentation in that package for more details.
+#'
+#' @param data_tb   a SITS tibble time series with an `matches` column. If `data_tb` is NULL, the function returns
+#' a pre-parameterized function to be called further in order to compute `multinom` training model
+#' @param formula   a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
+#' @param ...       other parameters to be passed to nnet::multinom function
+#' @return result   either an nnet::multinom class or an function prepared that can be called further to compute multinom training model
+#' @export
+#'
+sits_glm <- function(data_tb = NULL, formula = sits_formula_logref(), ...) {
+
+    # function that returns nnet::multinom model based on a sits sample tibble
+    result_fun <- function(tb){
+
+        # if parameter formula is a function call it passing as argument the input data sample. The function must return a valid formula.
+        if (class(formula) == "function")
+            formula <- formula(tb)
+
+        # call nnet::multinom method and return the trained multinom model
+        result_mlr <- nnet::multinom(formula = formula, data = tb, ...)
+        return(result_mlr)
+    }
+
+    # if no data is given, we prepare a function to compute multinom as model
+    if (is.null(data_tb))
+        result <- result_fun
+    # ...otherwise compute multinom model and return it
+    else
+        result <- result_fun(data_tb)
+    return(result)
+}
+
 #' @title Train SITS classifiction models
 #' @name sits_formula_logref
 #'
@@ -192,21 +318,21 @@ sits_formula_logref <- function(predictors_index = NULL){
 #' @description Given a SITS tibble time series and a model trained by \code{\link[sits]{sits_train_svm}},
 #' returns a SITS tibble with the classification.
 #'
-#' @param data.tb a SITS tibble time series
+#' @param data_tb a SITS tibble time series
 #' @param model a model trained by \code{\link[sits]{sits_train_svm}}
 #' @param ... other parameters to pass to \code{\link[sits]{sits_patterns}} and
 #' \code{\link[sits]{sits_TWDTW_matches}}
 #'
 #' @export
-sits_predict <- function(data.tb = NULL, model){
+sits_predict <- function(data_tb = NULL, model){
 
     # is the input data the result of a TWDTW matching function?
-    ensurer::ensure_that(data.tb, "matches" %in% names (.), err_desc = "sits_train_svm: input data does not contain TWDTW matches")
+    ensurer::ensure_that(data_tb, "matches" %in% names (.), err_desc = "sits_train_svm: input data does not contain TWDTW matches")
 
     # Spread TWDTW matches
-    spread.tb <- .sits_spread_matches(data.tb)
+    spread_tb <- .sits_spread_matches(data_tb)
 
-    data.tb$predicted <- as.character(stats::predict(model, newdata = spread.tb))
+    data_tb$predicted <- as.character(stats::predict(model, newdata = spread_tb))
 
-    return(data.tb)
+    return(data_tb)
 }
diff --git a/R/sits_table.R b/R/sits_table.R
index 3ea347d76..5581b2406 100644
--- a/R/sits_table.R
+++ b/R/sits_table.R
@@ -749,6 +749,7 @@ sits_relabel_conv <- function (file = NULL){
 #' @author Victor Maus, \email{vwmaus1@@gmail.com}
 #' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
 #' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
 #'
 #' @description Given a SITS tibble with a matches, returns a tibble whose columns have
 #' the reference label and the TWDTW distances for each temporal pattern.

From f4fa2441849287b614576d4f397ca1acf086e671 Mon Sep 17 00:00:00 2001
From: Rolf Simoes <rolfsimoes@gmail.com>
Date: Fri, 18 Aug 2017 12:45:17 -0300
Subject: [PATCH 2/7] Update ML functions

---
 R/sits_machine_learning.R | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/R/sits_machine_learning.R b/R/sits_machine_learning.R
index be7d11a04..724f98092 100644
--- a/R/sits_machine_learning.R
+++ b/R/sits_machine_learning.R
@@ -180,21 +180,21 @@ sits_mlr <- function(data.tb = NULL, formula = sits_formula_logref(), ...) {
 #' @param data.tb   a SITS tibble time series with an `matches` column. If `data.tb` is NULL, the function returns
 #' a pre-parameterized function to be called further in order to compute `multinom` training model
 #' @param formula   a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
-#' @param ...       other parameters to be passed to nnet::multinom function
-#' @return result   either an nnet::multinom class or an function prepared that can be called further to compute multinom training model
+#' @param ...       other parameters to be passed to glmnet::multinom function
+#' @return result   either an glmnet::multinom class or an function prepared that can be called further to compute multinom training model
 #' @export
 #'
 sits_glm <- function(data.tb = NULL, formula = sits_formula_logref(), ...) {
 
-    # function that returns nnet::multinom model based on a sits sample tibble
+    # function that returns glmnet::multinom model based on a sits sample tibble
     result_fun <- function(tb){
 
         # if parameter formula is a function call it passing as argument the input data sample. The function must return a valid formula.
         if (class(formula) == "function")
             formula <- formula(tb)
 
-        # call nnet::multinom method and return the trained multinom model
-        result_mlr <- nnet::multinom(formula = formula, data = tb, ...)
+        # call glmnet::multinom method and return the trained multinom model
+        result_mlr <- glmnet::glmnet(formula = formula, data = tb, ...)
         return(result_mlr)
     }
 

From ec8752edb0bfb5775cc9d40d8ff84674f8116180 Mon Sep 17 00:00:00 2001
From: Rolf Simoes <rolfsimoes@gmail.com>
Date: Fri, 18 Aug 2017 15:33:28 -0300
Subject: [PATCH 3/7] New ML functions

---
 DESCRIPTION                |  4 ++--
 NAMESPACE                  |  3 +++
 R/sits_assessment.R        | 13 ++++++++-----
 R/sits_init.R              |  4 +++-
 R/sits_machine_learning.R  |  2 +-
 conf_matrix.json           |  1 -
 man/sits_glm.Rd            | 33 +++++++++++++++++++++++++++++++++
 man/sits_lda.Rd            | 33 +++++++++++++++++++++++++++++++++
 man/sits_mlr.Rd            | 33 +++++++++++++++++++++++++++++++++
 man/sits_spread_matches.Rd |  2 ++
 man/sits_svm.Rd            |  4 ++--
 11 files changed, 120 insertions(+), 12 deletions(-)
 delete mode 100644 conf_matrix.json
 create mode 100644 man/sits_glm.Rd
 create mode 100644 man/sits_lda.Rd
 create mode 100644 man/sits_mlr.Rd

diff --git a/DESCRIPTION b/DESCRIPTION
index 85c8663fc..764066d84 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,8 +1,8 @@
 Package: sits
 Type: Package
 Title: Satellite Image Time Series Analysis
-Version: 0.8.16
-Date: 2017-08-16
+Version: 0.8.18
+Date: 2017-08-18
 Authors@R: c(person('Gilberto', 'Camara', role = c('aut', 'cre'), email = 'gilberto.camara@inpe.br'), 
              person('Rolf', 'Simoes', role = c('aut'), email = 'rolf.simoes@inpe.br'),
              person('Victor', 'Maus', role = c('aut'), email = 'vwmaus1@gmail.com'),
diff --git a/NAMESPACE b/NAMESPACE
index ad0663d9a..335c88316 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -22,6 +22,7 @@ export(sits_formula_logref)
 export(sits_gam)
 export(sits_getcovWTSS)
 export(sits_getdata)
+export(sits_glm)
 export(sits_group_bylatlong)
 export(sits_infoWTSS)
 export(sits_interp)
@@ -31,10 +32,12 @@ export(sits_kohonen)
 export(sits_labels)
 export(sits_labels_sample)
 export(sits_lag_diff)
+export(sits_lda)
 export(sits_linear_interp)
 export(sits_max_colors)
 export(sits_merge)
 export(sits_missing_values)
+export(sits_mlr)
 export(sits_mutate)
 export(sits_patterns)
 export(sits_plot)
diff --git a/R/sits_assessment.R b/R/sits_assessment.R
index d89b645f3..975bed69d 100644
--- a/R/sits_assessment.R
+++ b/R/sits_assessment.R
@@ -40,11 +40,14 @@ sits_accuracy <- function(pred.vec, ref.vec, pred_sans_ext = FALSE, conv.lst = N
     }
 
     if (NCOL(conf.mtx) != NROW (conf.mtx)) {
-        missing_names = colnames (conf.mtx) [!(colnames(conf.mtx) %in% row.names(conf.mtx))]
-        for (i in 1:length (missing_names)) {
-            vz <- rep (0, NCOL(conf.mtx))
-            conf.mtx <- rbind (conf.mtx, missing_names[i] = vz)
-        }
+        new.mtx <- matrix(rep(0, NROW(conf.mtx) * NCOL(conf.mtx)), nrow = NROW(conf.mtx), ncol = NCOL(conf.mtx))
+        new.mtx[colnames(conf.mtx) %in% row.names(conf.mtx), ] <- conf.mtx
+        # missing_names = colnames (conf.mtx) [!(colnames(conf.mtx) %in% row.names(conf.mtx))]
+        # for (i in 1:length (missing_names)) {
+        #     vz <- rep (0, NCOL(conf.mtx))
+        #     conf.mtx <- rbind (conf.mtx, missing_names[i] = vz)
+        # }
+        conf.mtx <- new.mtx
     }
     # ensures that the confusion matrix is square
     ensurer::ensure_that(conf.mtx, NCOL(.) == NROW(.),
diff --git a/R/sits_init.R b/R/sits_init.R
index f17747604..eca8450fe 100644
--- a/R/sits_init.R
+++ b/R/sits_init.R
@@ -1,8 +1,10 @@
+# On load
 .onAttach = function(lib, pkg){
      packageStartupMessage(
           sprintf("Loaded sits v%s. See ?sits for help, citation(\"sits\") for use in publication.\n",
                   utils::packageDescription("sits")$Version) )
 }
+
 # Include the following global variables in the SITS package
 #
 utils::globalVariables(c(".", "%>%", "Index", "value", "variable", "cond",
@@ -18,6 +20,6 @@ utils::globalVariables(c(".", "%>%", "Index", "value", "variable", "cond",
 #' @import dtwSat
 #' @importFrom magrittr %>%
 #' @usage lhs \%>\% rhs
-#'
 #' @useDynLib sits, .registration = TRUE
+#'
 NULL
diff --git a/R/sits_machine_learning.R b/R/sits_machine_learning.R
index 3cbec9ac5..e573f40fa 100644
--- a/R/sits_machine_learning.R
+++ b/R/sits_machine_learning.R
@@ -263,7 +263,7 @@ sits_predict <- function(data.tb = NULL, model){
     # Spread TWDTW matches
     spread.tb <- sits_spread_matches(data.tb)
 
-    data.tb$predicted <- as.character(stats::predict(model, newdata = spread_tb))
+    data.tb$predicted <- as.character(stats::predict(model, newdata = spread.tb))
 
     return(data.tb)
 }
diff --git a/conf_matrix.json b/conf_matrix.json
deleted file mode 100644
index 811595c70..000000000
--- a/conf_matrix.json
+++ /dev/null
@@ -1 +0,0 @@
-["Pasture", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Cerrado", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Pasture", "Pasture", "Cerrado", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Cerrado", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture", "Pasture"]
diff --git a/man/sits_glm.Rd b/man/sits_glm.Rd
new file mode 100644
index 000000000..74fd72f42
--- /dev/null
+++ b/man/sits_glm.Rd
@@ -0,0 +1,33 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/sits_machine_learning.R
+\name{sits_glm}
+\alias{sits_glm}
+\title{Train SITS classifiction models}
+\usage{
+sits_glm(data.tb = NULL, formula = sits_formula_logref(), ...)
+}
+\arguments{
+\item{data.tb}{a SITS tibble time series with an `matches` column. If `data.tb` is NULL, the function returns
+a pre-parameterized function to be called further in order to compute `multinom` training model}
+
+\item{formula}{a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)}
+
+\item{...}{other parameters to be passed to glmnet::multinom function}
+}
+\value{
+result   either an glmnet::multinom class or an function prepared that can be called further to compute multinom training model
+}
+\description{
+Use generalized liner model (glm) via penalized maximim likelihood to classify data.
+This function is a front-end to the "glmnet" method in the "glmnet" package.
+Please refer to the documentation in that package for more details.
+}
+\author{
+Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@ipea.gov.br}
+
+Victor Maus, \email{vwmaus1@gmail.com}
+
+Gilberto Camara, \email{gilberto.camara@inpe.br}
+
+Rolf Simoes, \email{rolf.simoes@inpe.br}
+}
diff --git a/man/sits_lda.Rd b/man/sits_lda.Rd
new file mode 100644
index 000000000..27a404ce2
--- /dev/null
+++ b/man/sits_lda.Rd
@@ -0,0 +1,33 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/sits_machine_learning.R
+\name{sits_lda}
+\alias{sits_lda}
+\title{Train SITS classifiction models}
+\usage{
+sits_lda(data.tb = NULL, formula = sits_formula_logref(), ...)
+}
+\arguments{
+\item{data.tb}{a SITS tibble time series with an `matches` column. If `data.tb` is NULL, the function returns
+a pre-parameterized function to be called further in order to compute `lda` training model}
+
+\item{formula}{a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)}
+
+\item{...}{other parameters to be passed to MASS::lda function}
+}
+\value{
+result   either an MASS::lda class or an function prepared that can be called further to compute lda training model
+}
+\description{
+Performs a linear discriminant analysis (lda) to classify data.
+This function is a front-end to the "lda" method in the "MASS" package.
+Please refer to the documentation in that package for more details.
+}
+\author{
+Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@ipea.gov.br}
+
+Victor Maus, \email{vwmaus1@gmail.com}
+
+Gilberto Camara, \email{gilberto.camara@inpe.br}
+
+Rolf Simoes, \email{rolf.simoes@inpe.br}
+}
diff --git a/man/sits_mlr.Rd b/man/sits_mlr.Rd
new file mode 100644
index 000000000..3586a65c8
--- /dev/null
+++ b/man/sits_mlr.Rd
@@ -0,0 +1,33 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/sits_machine_learning.R
+\name{sits_mlr}
+\alias{sits_mlr}
+\title{Train SITS classifiction models}
+\usage{
+sits_mlr(data.tb = NULL, formula = sits_formula_logref(), ...)
+}
+\arguments{
+\item{data.tb}{a SITS tibble time series with an `matches` column. If `data.tb` is NULL, the function returns
+a pre-parameterized function to be called further in order to compute `multinom` training model}
+
+\item{formula}{a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)}
+
+\item{...}{other parameters to be passed to nnet::multinom function}
+}
+\value{
+result   either an nnet::multinom class or an function prepared that can be called further to compute multinom training model
+}
+\description{
+Use multinomial log-linear (mlr) fitting model via neural networks to classify data.
+This function is a front-end to the "multinom" method in the "nnet" package.
+Please refer to the documentation in that package for more details.
+}
+\author{
+Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@ipea.gov.br}
+
+Victor Maus, \email{vwmaus1@gmail.com}
+
+Gilberto Camara, \email{gilberto.camara@inpe.br}
+
+Rolf Simoes, \email{rolf.simoes@inpe.br}
+}
diff --git a/man/sits_spread_matches.Rd b/man/sits_spread_matches.Rd
index f52a47478..b4e3e6be5 100644
--- a/man/sits_spread_matches.Rd
+++ b/man/sits_spread_matches.Rd
@@ -22,4 +22,6 @@ Victor Maus, \email{vwmaus1@gmail.com}
 Gilberto Camara, \email{gilberto.camara@inpe.br}
 
 Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@ipea.gov.br}
+
+Rolf Simoes, \email{rolf.simoes@inpe.br}
 }
diff --git a/man/sits_svm.Rd b/man/sits_svm.Rd
index bc4601077..328c1563e 100644
--- a/man/sits_svm.Rd
+++ b/man/sits_svm.Rd
@@ -9,10 +9,10 @@ sits_svm(data.tb = NULL, formula = sits_formula_logref(),
   tolerance = 0.001, epsilon = 0.1, ...)
 }
 \arguments{
-\item{data.tb}{a SITS tibble time series with an alignment column. If data_tb is NULL, the function returns
+\item{data.tb}{a SITS tibble time series with an alignment column. If data.tb is NULL, the function returns
 a function to be called further to compute svm training model according to given parameters}
 
-\item{formula}{a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_svm)}
+\item{formula}{a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)}
 
 \item{kernel}{the kernel used in training and predicting (options = linear, polynomial, radial basis, sigmoid)}
 

From cb5111c307a421bacf8390e1f6fa67fa6db799a7 Mon Sep 17 00:00:00 2001
From: Rolf Simoes <rolfsimoes@gmail.com>
Date: Fri, 18 Aug 2017 19:13:15 -0300
Subject: [PATCH 4/7] Improving package cpompilation

---
 inst/extdata/examples/sits_test_machine_learning.R | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/inst/extdata/examples/sits_test_machine_learning.R b/inst/extdata/examples/sits_test_machine_learning.R
index d2e251da8..bd7075dc5 100644
--- a/inst/extdata/examples/sits_test_machine_learning.R
+++ b/inst/extdata/examples/sits_test_machine_learning.R
@@ -16,4 +16,4 @@ predict.tb <- sits_predict(matches.tb, obj.svm)
 
 sits_accuracy(ref.vec = predict.tb$label, pred.vec = predict.tb$predicted)
 
-sits_kfold_validate(cerrado.tb, folds = 3)
+sits_kfold_validate(cerrado.tb, folds = 2)

From 8030649260cb91300f2746fc13bdba3d069eccf9 Mon Sep 17 00:00:00 2001
From: Rolf Simoes <rolfsimoes@gmail.com>
Date: Mon, 21 Aug 2017 15:39:25 -0300
Subject: [PATCH 5/7] New sits cluster functions

---
 DESCRIPTION                   |   4 +-
 NAMESPACE                     |   4 ++
 R/sits_TWDTW.R                |   2 +-
 R/sits_cluster.R              | 125 +++++++++++++++++++++++++++++++++-
 man/sits_cluster.Rd           |   4 +-
 man/sits_cluster_cleaner.Rd   |  23 +++++++
 man/sits_cluster_frequency.Rd |  23 +++++++
 man/sits_cluster_relabel.Rd   |  24 +++++++
 man/sits_cluster_validity.Rd  |  27 ++++++++
 9 files changed, 227 insertions(+), 9 deletions(-)
 create mode 100644 man/sits_cluster_cleaner.Rd
 create mode 100644 man/sits_cluster_frequency.Rd
 create mode 100644 man/sits_cluster_relabel.Rd
 create mode 100644 man/sits_cluster_validity.Rd

diff --git a/DESCRIPTION b/DESCRIPTION
index 764066d84..f80d8473c 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,8 +1,8 @@
 Package: sits
 Type: Package
 Title: Satellite Image Time Series Analysis
-Version: 0.8.18
-Date: 2017-08-18
+Version: 0.8.21
+Date: 2017-08-21
 Authors@R: c(person('Gilberto', 'Camara', role = c('aut', 'cre'), email = 'gilberto.camara@inpe.br'), 
              person('Rolf', 'Simoes', role = c('aut'), email = 'rolf.simoes@inpe.br'),
              person('Victor', 'Maus', role = c('aut'), email = 'vwmaus1@gmail.com'),
diff --git a/NAMESPACE b/NAMESPACE
index 17fad51fc..3dfe7a0e9 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -8,6 +8,10 @@ export(sits_align)
 export(sits_apply)
 export(sits_bands)
 export(sits_cluster)
+export(sits_cluster_cleaner)
+export(sits_cluster_frequency)
+export(sits_cluster_relabel)
+export(sits_cluster_validity)
 export(sits_color_name)
 export(sits_coverageWTSS)
 export(sits_createColRowSequence)
diff --git a/R/sits_TWDTW.R b/R/sits_TWDTW.R
index a3048c80d..901adff1a 100644
--- a/R/sits_TWDTW.R
+++ b/R/sits_TWDTW.R
@@ -48,7 +48,7 @@ sits_TWDTW_matches <- function (data.tb = NULL, patterns.tb = NULL, bands = NULL
         .sits_test_table (patterns.tb)
 
         # handle the case of null bands
-        if (purrr::is_null (bands)) bands <- sits_bands(tb)
+        if (purrr::is_null (bands)) bands <- sits_bands(data.tb)
 
         # create a tibble to store the results of the TWDTW matches
         matches.tb <- sits_table()
diff --git a/R/sits_cluster.R b/R/sits_cluster.R
index 9d1bbe1ca..108bf37f5 100644
--- a/R/sits_cluster.R
+++ b/R/sits_cluster.R
@@ -10,17 +10,136 @@
 #' @param data.tb          a tibble with input data of dtwclust.
 #' @param clusters         a cluster structure returned from dtwclust.
 #' @param k                the desired number of clusters
-#' @param height           the height of dendrogram cutting
 #' @return result.tb       a SITS table with the clusters or clusters' members
 #' @export
-sits_cluster <-  function (data.tb, clusters, k = NULL, height = NULL) {
+sits_cluster <-  function (data.tb, clusters, k) {
 
     # create a table to store the results
     result.tb <- data.tb
 
     # cut the tree
-    result.tb$cluster <- stats::cutree(clusters, k, height)
+    result.tb$cluster <- stats::cutree(clusters, k)
 
     return (result.tb)
 }
 
+#' @title Cluster validity indices
+#' @name sits_cluster_validity
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Compute different cluster validity indices. This function needs
+#' as input a SITS tibble with `cluster` column. It is a front-end to
+#' `dtwclust::cvi` function. Please refer to the documentation in that package for more details.
+#' @references "dtwclust" package (https://CRAN.R-project.org/package=dtwclust)
+#'
+#' @param data.tb          a SITS tibble with `cluster` column.
+#' @param type             character vector indicating which indices are to be computed. (Default "valid")
+#' @return result.vec      vectors with chosen CVIs
+#' @export
+sits_cluster_validity <-  function (data.tb, type = "valid") {
+
+    # is the input data the result of a cluster function?
+    ensurer::ensure_that(data.tb, "cluster" %in% names (.), err_desc = "sits_cluster_cleaner: input data does not contain cluster column")
+
+    # compute CVIs and return
+    result.vec <- dtwclust::cvi(a = data.tb$cluster, b = factor(data.tb$label), type = type, log.base = 10)
+
+    return (result.vec)
+}
+
+#' @title Cluster frequency
+#' @name sits_cluster_frequency
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Computes the frequency of labels in each cluster.
+#' This function needs as input a SITS tibble with `cluster` column.
+#'
+#' @param data.tb          a SITS tibble with `cluster` column.
+#' @param relative         (boolean) return relative frequency?
+#' @return result.mtx      matrix containing all frequencies of labels in clusters
+#' @export
+sits_cluster_frequency <-  function (data.tb, relative = FALSE) {
+
+    # is the input data the result of a cluster function?
+    ensurer::ensure_that(data.tb, "cluster" %in% names (.), err_desc = "sits_cluster_cleaner: input data does not contain cluster column")
+
+    # compute frequency table
+    result.mtx <- table(data.tb$label, data.tb$cluster)
+
+    # compute relative frequency
+    if (relative)
+        result.mtx <- prop.table(result.mtx, margin = 2)
+
+    # compute total row and col
+    result.mtx <- stats::addmargins(result.mtx, FUN = list(Total = sum), quiet = TRUE)
+    return (result.mtx)
+}
+
+#' @title Cluster cleaner
+#' @name sits_cluster_cleaner
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Removes SITS tibble samples of labels that are minority in each cluster.
+#' This function needs as input a SITS tibble with `cluster` column.
+#'
+#' @param data.tb          a SITS tibble with `cluster` column.
+#' @param min_clu_perc     minimum percentage of representativeness inside a cluster to remain in cluster.
+#' @return result.tb       a SITS tibble with all selected samples
+#' @export
+sits_cluster_cleaner <-  function (data.tb, min_clu_perc) {
+
+    # is the input data the result of a cluster function?
+    ensurer::ensure_that(data.tb, "cluster" %in% names (.), err_desc = "sits_cluster_cleaner: input data does not contain cluster column")
+
+    # compute frequency in each cluster
+    freq.mtx <- sits_cluster_frequency(data.tb, relative = TRUE)
+
+    # get those indexes whose labels represents more than `min_clu_perc`
+    index.mtx <- which(freq.mtx[1:NROW(freq.mtx) - 1,1:NCOL(freq.mtx) - 1] > min_clu_perc, arr.ind = TRUE, useNames = TRUE)
+
+    # return only those samples that satisfies the `min_clu_perc` condition
+    filter_condition <- paste0(purrr::map2(rownames(index.mtx), index.mtx[,2],
+                                           function(lbl, clu) paste0("label=='", lbl, "' & cluster==", clu)),
+                               collapse = " | ")
+
+    # if no index selescted, return none
+    filter_condition <- ifelse(filter_condition != "", filter_condition, "FALSE")
+
+    # filter result and return
+    result.tb <- dplyr::filter_(data.tb, filter_condition)
+    return (result.tb)
+}
+
+#' @title Cluster cleaner
+#' @name sits_cluster_relabel
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Renames the labels of SITS tibble samples according to its respective cluster.
+#' This function needs as input a SITS tibble with `cluster` column.
+#'
+#' @param data.tb          a SITS tibble with `cluster` column.
+#' @param cluster_names    character vector informing all cluster names. If unnamed vector is informed,
+#'                         the index of each name will be treated as cluster code
+#' @return result.tb       SITS tibble with relabeled samples
+#' @export
+sits_cluster_relabel <-  function (data.tb, cluster_names) {
+
+    # is the input data the result of a cluster function?
+    ensurer::ensure_that(data.tb, "cluster" %in% names (.), err_desc = "sits_cluster_cleaner: input data does not contain cluster column")
+
+    # compute new clusters names
+    # if an unnamed vector is given
+    if (is.null(names(cluster_names)))
+        data_cluster_names.vec <- cluster_names[data.tb$cluster]
+    # if a named vector is given
+    else
+        data_cluster_names.vec <- cluster_names[as.character(data.tb$cluster)]
+
+    # fill not renamed entries with original value
+    data_cluster_names.vec <- ifelse(is.na(data_cluster_names.vec), data.tb$cluster, data_cluster_names.vec)
+    names(data_cluster_names.vec) <- NULL
+
+    # relabel result and return
+    result.tb$label <- data_cluster_names.vec
+    return (result.tb)
+}
diff --git a/man/sits_cluster.Rd b/man/sits_cluster.Rd
index ba07647c2..d7cca9f3b 100644
--- a/man/sits_cluster.Rd
+++ b/man/sits_cluster.Rd
@@ -4,7 +4,7 @@
 \alias{sits_cluster}
 \title{Cuts a cluster tree produced by sits_dendrogram}
 \usage{
-sits_cluster(data.tb, clusters, k = NULL, height = NULL)
+sits_cluster(data.tb, clusters, k)
 }
 \arguments{
 \item{data.tb}{a tibble with input data of dtwclust.}
@@ -12,8 +12,6 @@ sits_cluster(data.tb, clusters, k = NULL, height = NULL)
 \item{clusters}{a cluster structure returned from dtwclust.}
 
 \item{k}{the desired number of clusters}
-
-\item{height}{the height of dendrogram cutting}
 }
 \value{
 result.tb       a SITS table with the clusters or clusters' members
diff --git a/man/sits_cluster_cleaner.Rd b/man/sits_cluster_cleaner.Rd
new file mode 100644
index 000000000..708fa0f8a
--- /dev/null
+++ b/man/sits_cluster_cleaner.Rd
@@ -0,0 +1,23 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/sits_cluster.R
+\name{sits_cluster_cleaner}
+\alias{sits_cluster_cleaner}
+\title{Cluster cleaner}
+\usage{
+sits_cluster_cleaner(data.tb, min_clu_perc)
+}
+\arguments{
+\item{data.tb}{a SITS tibble with `cluster` column.}
+
+\item{min_clu_perc}{minimum percentage of representativeness inside a cluster to remain in cluster.}
+}
+\value{
+result.tb       a SITS tibble with all selected samples
+}
+\description{
+Removes SITS tibble samples of labels that are minority in each cluster.
+This function needs as input a SITS tibble with `cluster` column.
+}
+\author{
+Rolf Simoes, \email{rolf.simoes@inpe.br}
+}
diff --git a/man/sits_cluster_frequency.Rd b/man/sits_cluster_frequency.Rd
new file mode 100644
index 000000000..542edf8de
--- /dev/null
+++ b/man/sits_cluster_frequency.Rd
@@ -0,0 +1,23 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/sits_cluster.R
+\name{sits_cluster_frequency}
+\alias{sits_cluster_frequency}
+\title{Cluster frequency}
+\usage{
+sits_cluster_frequency(data.tb, relative = FALSE)
+}
+\arguments{
+\item{data.tb}{a SITS tibble with `cluster` column.}
+
+\item{relative}{(boolean) return relative frequency?}
+}
+\value{
+result.mtx      matrix containing all frequencies of labels in clusters
+}
+\description{
+Computes the frequency of labels in each cluster.
+This function needs as input a SITS tibble with `cluster` column.
+}
+\author{
+Rolf Simoes, \email{rolf.simoes@inpe.br}
+}
diff --git a/man/sits_cluster_relabel.Rd b/man/sits_cluster_relabel.Rd
new file mode 100644
index 000000000..15b2278d0
--- /dev/null
+++ b/man/sits_cluster_relabel.Rd
@@ -0,0 +1,24 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/sits_cluster.R
+\name{sits_cluster_relabel}
+\alias{sits_cluster_relabel}
+\title{Cluster cleaner}
+\usage{
+sits_cluster_relabel(data.tb, cluster_names)
+}
+\arguments{
+\item{data.tb}{a SITS tibble with `cluster` column.}
+
+\item{cluster_names}{character vector informing all cluster names. If unnamed vector is informed,
+the index of each name will be treated as cluster code}
+}
+\value{
+result.tb       SITS tibble with relabeled samples
+}
+\description{
+Renames the labels of SITS tibble samples according to its respective cluster.
+This function needs as input a SITS tibble with `cluster` column.
+}
+\author{
+Rolf Simoes, \email{rolf.simoes@inpe.br}
+}
diff --git a/man/sits_cluster_validity.Rd b/man/sits_cluster_validity.Rd
new file mode 100644
index 000000000..e78ea87dc
--- /dev/null
+++ b/man/sits_cluster_validity.Rd
@@ -0,0 +1,27 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/sits_cluster.R
+\name{sits_cluster_validity}
+\alias{sits_cluster_validity}
+\title{Cluster validity indices}
+\usage{
+sits_cluster_validity(data.tb, type = "valid")
+}
+\arguments{
+\item{data.tb}{a SITS tibble with `cluster` column.}
+
+\item{type}{character vector indicating which indices are to be computed. (Default "valid")}
+}
+\value{
+result.vec      vectors with chosen CVIs
+}
+\description{
+Compute different cluster validity indices. This function needs
+as input a SITS tibble with `cluster` column. It is a front-end to
+`dtwclust::cvi` function. Please refer to the documentation in that package for more details.
+}
+\references{
+"dtwclust" package (https://CRAN.R-project.org/package=dtwclust)
+}
+\author{
+Rolf Simoes, \email{rolf.simoes@inpe.br}
+}

From 9c6d17e668e4e39d25f6bae8f247e25c6bd7259b Mon Sep 17 00:00:00 2001
From: Rolf Simoes <rolfsimoes@gmail.com>
Date: Mon, 21 Aug 2017 17:44:12 -0300
Subject: [PATCH 6/7] Fix sits_cluster_relabel bug

---
 R/sits_cluster.R | 1 +
 1 file changed, 1 insertion(+)

diff --git a/R/sits_cluster.R b/R/sits_cluster.R
index 108bf37f5..907bca7b0 100644
--- a/R/sits_cluster.R
+++ b/R/sits_cluster.R
@@ -140,6 +140,7 @@ sits_cluster_relabel <-  function (data.tb, cluster_names) {
     names(data_cluster_names.vec) <- NULL
 
     # relabel result and return
+    result.tb <- data.tb
     result.tb$label <- data_cluster_names.vec
     return (result.tb)
 }

From 0359c42fc154337a648e1d47a297ba778220a361 Mon Sep 17 00:00:00 2001
From: Rolf Simoes <rolfsimoes@gmail.com>
Date: Tue, 22 Aug 2017 18:28:59 -0300
Subject: [PATCH 7/7] Add new Machine Learn functions Change sits_predict
 function Change model data structure returned by sits_train

---
 R/sits_machine_learning.R                     | 204 +++++++++++++-
 .../Modelagem_classificacao_dados_dtw_v3.R    | 250 +++++++++---------
 2 files changed, 320 insertions(+), 134 deletions(-)

diff --git a/R/sits_machine_learning.R b/R/sits_machine_learning.R
index 4a9c11490..8437a7ede 100644
--- a/R/sits_machine_learning.R
+++ b/R/sits_machine_learning.R
@@ -60,7 +60,7 @@ sits_train <- function(distances.tb, tr_method = sits_svm()){
 #' @param tolerance	       tolerance of termination criterion (default: 0.001)
 #' @param epsilon	       epsilon in the insensitive-loss function (default: 0.1)
 #' @param ...              other parameters to be passed to e1071::svm function
-#' @return result          either an e1071::svm class or an function prepared that can be called further to compute svm training model
+#' @return result          either an model function to be passed in sits_predict or an function prepared that can be called further to compute multinom training model
 #' @export
 #'
 sits_svm <- function(distances.tb = NULL, formula = sits_formula_logref(), kernel = "linear",
@@ -79,10 +79,194 @@ sits_svm <- function(distances.tb = NULL, formula = sits_formula_logref(), kerne
         # call e1071::svm method and return the trained svm model
         result_svm <- e1071::svm(formula = formula, data = tb, kernel = kernel,
                                  degree = degree, cost = cost, coef0 = coef0, tolerance = tolerance, epsilon = epsilon, ...)
-        return(result_svm)
+
+        # construct model predict enclosure function and returns
+        model_predict <- function(test_distances.tb){
+            return(stats::predict(result_lda, newdata = test_distances.tb))
+        }
+        return(model_predict)
+    }
+
+    result <- .sits_factory_function (distances.tb, result_fun)
+}
+#' @title Train SITS classifiction models
+#' @name sits_lda
+#'
+#' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Victor Maus, \email{vwmaus1@@gmail.com}
+#' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Performs a linear discriminant analysis (lda) to classify data.
+#' This function is a front-end to the "lda" method in the "MASS" package.
+#' Please refer to the documentation in that package for more details.
+#'
+#' @param distances.tb     a time series with a set of distance measures for each training sample
+#' @param formula          a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
+#' @param ...              other parameters to be passed to MASS::lda function
+#' @return result          either an model function to be passed in sits_predict or an function prepared that can be called further to compute multinom training model
+#' @export
+#'
+sits_lda <- function(distances.tb = NULL, formula = sits_formula_logref(), ...) {
+
+    # function that returns MASS::lda model based on a sits sample tibble
+    result_fun <- function(train_distances.tb){
+
+        # is the input data the result of a TWDTW matching function?
+        ensurer::ensure_that(train_distances.tb, "reference" %in% names (.), err_desc = "sits_lda: input data does not contain distance")
+
+        # if parameter formula is a function call it passing as argument the input data sample. The function must return a valid formula.
+        if (class(formula) == "function")
+            formula <- formula(train_distances.tb)
+
+        # call MASS::lda method and return the trained lda model
+        result_lda <- MASS::lda(formula = formula, data = train_distances.tb, ...)
+
+        # construct model predict enclosure function and returns
+        model_predict <- function(test_distances.tb){
+            return(stats::predict(result_lda, newdata = test_distances.tb))
+        }
+        return(model_predict)
+    }
+
+    result <- .sits_factory_function (distances.tb, result_fun)
+    return(result)
+}
+
+#' @title Train SITS classifiction models
+#' @name sits_mlr
+#'
+#' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Victor Maus, \email{vwmaus1@@gmail.com}
+#' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Use multinomial log-linear (mlr) fitting model via neural networks to classify data.
+#' This function is a front-end to the "multinom" method in the "nnet" package.
+#' Please refer to the documentation in that package for more details.
+#'
+#' @param distances.tb     a time series with a set of distance measures for each training sample
+#' @param formula          a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
+#' @param ...              other parameters to be passed to nnet::multinom function
+#' @return result          either an model function to be passed in sits_predict or an function prepared that can be called further to compute multinom training model
+#' @export
+#'
+sits_mlr <- function(distances.tb = NULL, formula = sits_formula_logref(), ...) {
+
+    # function that returns nnet::multinom model based on a sits sample tibble
+    result_fun <- function(train_distances.tb){
+
+        # is the input data the result of a TWDTW matching function?
+        ensurer::ensure_that(train_distances.tb, "reference" %in% names (.), err_desc = "sits_mlr: input data does not contain distance")
+
+        # if parameter formula is a function call it passing as argument the input data sample. The function must return a valid formula.
+        if (class(formula) == "function")
+            formula <- formula(train_distances.tb)
+
+        # call nnet::multinom method and return the trained multinom model
+        result_mlr <- nnet::multinom(formula = formula, data = train_distances.tb, ...)
+
+        # construct model predict enclosure function and returns
+        model_predict <- function(test_distances.tb){
+            return(stats::predict(result_mlr, newdata = test_distances.tb))
+        }
+        return(model_predict)
+    }
+
+    result <- .sits_factory_function (distances.tb, result_fun)
+    return(result)
+}
+
+#' @title Train SITS classifiction models
+#' @name sits_glm
+#'
+#' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Victor Maus, \email{vwmaus1@@gmail.com}
+#' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Use generalized liner model (glm) via penalized maximim likelihood to classify data.
+#' This function is a front-end to the "cv.glmnet" method in the "glmnet" package.
+#' Please refer to the documentation in that package for more details.
+#'
+#' @param distances.tb     a time series with a set of distance measures for each training sample
+#' @param family           Response type. Can be either "gaussian", "binomial", "poisson", "multinomial", "cox", or "mgaussian". (default: "multinomial")
+#' @param alpha            the elasticnet mixing parameter, with 0<=alpha<=1. Set alpha = 0.0 to obtain ridge model, and alpha = 1.0 to obtain lasso model).
+#'                         (refer to `glmnet::cv.glmnet` function for more details)
+#' @param lambda_kfolds    number of folds to find best lambda parameter (default: 10)
+#' @param ...              other parameters to be passed to `glmnet::cv.glmnet` function
+#' @return result          either an model function to be passed in sits_predict or an function prepared that can be called further to compute multinom training model
+#' @export
+#'
+sits_glm <- function(distances.tb = NULL, family = "multinomial", alpha = 1.0, lambda_kfolds = 10, ...) {
+
+    # function that returns glmnet::multinom model based on a sits sample tibble
+    result_fun <- function(train_distances.tb){
+
+        # is the input data the result of a TWDTW matching function?
+        ensurer::ensure_that(train_distances.tb, "reference" %in% names (.), err_desc = "sits_glm: input data does not contain distance")
+
+        # call glmnet::multinom method and return the trained multinom model
+        result_glm <- glmnet::cv.glmnet(y = factor(data.matrix(train_distances.tb$reference)),
+                                        x = log(data.matrix(train_distances.tb[,3:NCOL(train_distances.tb)])),
+                                        family = family, alpha = alpha, k = lambda_kfolds, ...)
+
+        # construct model predict enclosure function and returns
+        model_predict <- function(test_distances.tb){
+            return(stats::predict(result_glm,
+                                  s = result_glm$lambda.min,
+                                  newx = log(data.matrix(test_distances.tb[,3:NCOL(test_distances.tb)])), type = "class"))
+        }
+        return(model_predict)
+    }
+
+    result <- .sits_factory_function (distances.tb, result_fun)
+    return(result)
+}
+
+#' @title Train SITS classifiction models
+#' @name sits_rfor
+#'
+#' @author Alexandre Xavier Ywata de Carvalho, \email{alexandre.ywata@@ipea.gov.br}
+#' @author Victor Maus, \email{vwmaus1@@gmail.com}
+#' @author Gilberto Camara, \email{gilberto.camara@@inpe.br}
+#' @author Rolf Simoes, \email{rolf.simoes@@inpe.br}
+#'
+#' @description Use Random Forest algorithm to classify data.
+#' This function is a front-end to the "randomForest" method in the "randomForest" package.
+#' Please refer to the documentation in that package for more details.
+#'
+#' @param distances.tb     a time series with a set of distance measures for each training sample
+#' @param n_tree           Number of trees to grow. This should not be set to too small a number,
+#'                         to ensure that every input row gets predicted at least a few times. (default: 500)
+#' @param node_size
+#' @param ...              other parameters to be passed to `randomForest::randomForest` function
+#' @return result          either an model function to be passed in sits_predict or an function prepared that can be called further to compute multinom training model
+#' @export
+#'
+sits_rfor <- function(distances.tb = NULL, n_tree = 500, ...) {
+
+    # function that returns `randomForest::randomForest` model based on a sits sample tibble
+    result_fun <- function(train_distances.tb){
+
+        # is the input data the result of a TWDTW matching function?
+        ensurer::ensure_that(train_distances.tb, "reference" %in% names (.), err_desc = "sits_mlr: input data does not contain distance")
+
+        # call `randomForest::randomForest` method and return the trained multinom model
+        result_rfor <- randomForest::randomForest(y = data.matrix(train_distances.tb$reference),
+                                                  x = log(data.matrix(train_distances.tb[,2:NCOL(train_distances.tb)])),
+                                                  data = NULL, ntree = n_tree, nodesize = 1,
+                                                  norm.votes = FALSE, train_distances.tb, ...)
+
+        # construct model predict enclosure function and returns
+        model_predict <- function(test_distances.tb){
+            return(stats::predict(result_rfor, newdata = test_distances.tb, type = "response"))
+        }
+        return(model_predict)
     }
 
     result <- .sits_factory_function (distances.tb, result_fun)
+    return(result)
 }
 
 #' @title Train SITS classifiction models
@@ -121,7 +305,6 @@ sits_formula_logref <- function(predictors_index = -2:0){
     return(result_fun)
 }
 
-
 #' @title Predict class based on the trained models
 #' @name sits_predict
 #'
@@ -133,18 +316,21 @@ sits_formula_logref <- function(predictors_index = -2:0){
 #' @description Given a SITS tibble time series and a model trained by \code{\link[sits]{sits_train}},
 #' returns a SITS tibble with the classification.
 #'
-#' @param data.tb a SITS tibble time series
-#' @param distances.tb a tibble with a set of distance metrics to each of the classes
-#' @param model a model trained by \code{\link[sits]{sits_train}}
-#' @return data.tb - a SITS tibble with the predicted label
+#' @param data.tb       a SITS tibble time series
+#' @param distances.tb  a tibble with a set of distance metrics to each of the classes
+#' @param model         a model trained by \code{\link[sits]{sits_train}}
+#' @return data.tb      a SITS tibble with the predicted label
 #'
 #' @export
-sits_predict <- function(data.tb = NULL, distances.tb = NULL, model){
+sits_predict <- function(data.tb = NULL, distances.tb = NULL, model, ...){
 
     # is the input data the result of a TWDTW matching function?
     ensurer::ensure_that(distances.tb, "reference" %in% names (.), err_desc = "sits_train_svm: input data does not contain TWDTW matches")
 
-    data.tb$predicted <- as.character(stats::predict(model, newdata = distances.tb))
+    # is the input model a model function?
+    ensurer::ensure_that(model, class (.) == "function", err_desc = "sits_predict: model parameter is not a function model returned by sits_train.")
+
+    data.tb$predicted <- as.character(model(distances.tb))
 
     return(data.tb)
 }
diff --git a/inst/extdata/ML/ProgramasR/Modelagem_classificacao_dados_dtw_v3.R b/inst/extdata/ML/ProgramasR/Modelagem_classificacao_dados_dtw_v3.R
index 4ef7d1301..14e9feb4a 100644
--- a/inst/extdata/ML/ProgramasR/Modelagem_classificacao_dados_dtw_v3.R
+++ b/inst/extdata/ML/ProgramasR/Modelagem_classificacao_dados_dtw_v3.R
@@ -25,13 +25,13 @@ rm(list=ls());
 #--- data used for estimations             ---#
 #---------------------------------------------#
 
-dir_base <- "D:\\Alex\\Pesquisa\\LandUseChange\\IIASA_GLOBIOM\\DTWMachineLearning"
+dir_base <- "~/sits/inst/extdata/ML"
 #dir_base <- "//storage3/usuarios/AlexandreYwata/Pesquisa/LandUseChange/IIASA_GLOBIOM/SITS/inst/extdata/examples/PosProcessamentoDTW"
 
-dir_dados <- paste0(dir_base, "\\Dados");
-dir_prog <- paste0(dir_base, "\\ProgramasR")
+dir_dados <- paste0(dir_base, "/Dados");
+dir_prog <- paste0(dir_base, "/ProgramasR")
 
-file_dados <- paste0(dir_dados, "\\mt_twdtw_distances.csv")
+file_dados <- paste0(dir_dados, "/mt_twdtw_distances.csv")
 
 dados_originais <- read.csv(file_dados, stringsAsFactors=FALSE, header=TRUE, sep = ",");
 str(dados_originais)
@@ -57,7 +57,7 @@ for (i in 1:nrow(dados))
     for (j in 1:length(list_refs))
     {
         variavel <- list_refs[j]
-        
+
         if (dados[i,variavel] < min_dist)
         {
             min_dist <- dados[i,variavel]
@@ -76,8 +76,8 @@ for (i in 1:nrow(dados))
 #               "Soybean_Comerc1" = "Double_Cropping",
 #               "Soybean_Comerc2" = "Double_Cropping",
 #               "Soybean_Cotton" = "Soybean_Cotton",
-#               "Soybean_Fallow1" = "Single_Cropping", 
-#               "Soybean_Fallow2" = "Single_Cropping", 
+#               "Soybean_Fallow1" = "Single_Cropping",
+#               "Soybean_Fallow2" = "Single_Cropping",
 #               "Soybean_NonComerc1" = "Double_Cropping",
 #               "Soybean_NonComerc2" = "Double_Cropping",
 #               "Soybean_Pasture" = "Pasture",
@@ -92,8 +92,8 @@ conv.lst <- c("Fallow_Cotton"  = "Cotton",
               "Soybean_Comerc1" = "Double_Cropping",
               "Soybean_Comerc2" = "Double_Cropping",
               "Soybean_Cotton" = "Double_Cropping",
-              "Soybean_Fallow1" = "Single_Cropping", 
-              "Soybean_Fallow2" = "Single_Cropping", 
+              "Soybean_Fallow1" = "Single_Cropping",
+              "Soybean_Fallow2" = "Single_Cropping",
               "Soybean_NonComerc1" = "Double_Cropping",
               "Soybean_NonComerc2" = "Double_Cropping",
               "Soybean_Pasture" = "Pasture",
@@ -146,10 +146,10 @@ table(dados$class, dados$classnum)
 #-----------------------------------------------------------#
 
 nomes <- names(dados); nomes
-lognomes <- paste0('log(', nomes[nomes %in% c(list_refs)], ')'); 
+lognomes <- paste0('log(', nomes[nomes %in% c(list_refs)], ')');
 paste(lognomes, collapse = " + ")
 
-orinomes <- paste0(nomes[nomes %in% c(list_refs)]); 
+orinomes <- paste0(nomes[nomes %in% c(list_refs)]);
 paste(orinomes, collapse = " + ")
 
 formula1 <- as.formula(paste("factor(class) ~ ", paste(lognomes, collapse = " + ")));
@@ -172,10 +172,10 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-        
+
     categorias.lda <- lda(formula1, data=dadosTrain)
     summary(categorias.lda)
-    
+
     categorias.lda.pred <- as.character(predict(categorias.lda, newdata = dadosTest)$class)
     dados[dados$folds == k, 'pred_lda'] <- categorias.lda.pred
 }
@@ -193,10 +193,10 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
+
     categorias.mlr <- multinom(formula1, data=dadosTrain)
     summary(categorias.mlr)
-    
+
     categorias.mlr.pred <- as.character(predict(categorias.mlr, newdata = dadosTest))
     dados[dados$folds == k, 'pred_mlr'] <- categorias.mlr.pred
 }
@@ -222,17 +222,17 @@ for (costsvm in list_costs)
         {
             dadosTrain <- dados[dados$folds != k,]
             dadosTest <- dados[dados$folds == k,]
-            
-            categorias.svm.linear <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                         kernel = "linear", 
-                                         type="C-classification", epsilon = epsilonsvm, cost = costsvm)     
+
+            categorias.svm.linear <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                         kernel = "linear",
+                                         type="C-classification", epsilon = epsilonsvm, cost = costsvm)
             summary(categorias.svm.linear)
-            
-            categorias.svm.linear.pred <- as.character(predict(categorias.svm.linear, 
+
+            categorias.svm.linear.pred <- as.character(predict(categorias.svm.linear,
                                                        newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
             dados[dados$folds == k, 'pred_svm_linear'] <- categorias.svm.linear.pred
         }
-        
+
         cv_svm_linear[counter, 1] <- costsvm
         cv_svm_linear[counter, 2] <- epsilonsvm
         cv_svm_linear[counter, 3] <- mean(dados$class == dados$pred_svm_linear)
@@ -249,13 +249,13 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
-    categorias.svm.linear <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                 kernel = "linear", 
-                                 type="C-classification", epsilon = epsilon_opt, cost = cost_opt)     
+
+    categorias.svm.linear <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                 kernel = "linear",
+                                 type="C-classification", epsilon = epsilon_opt, cost = cost_opt)
     summary(categorias.svm.linear)
-    
-    categorias.svm.linear.pred <- as.character(predict(categorias.svm.linear, 
+
+    categorias.svm.linear.pred <- as.character(predict(categorias.svm.linear,
                                                        newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
     dados[dados$folds == k, 'pred_svm_linear'] <- categorias.svm.linear.pred
 }
@@ -281,17 +281,17 @@ for (costsvm in list_costs)
         {
             dadosTrain <- dados[dados$folds != k,]
             dadosTest <- dados[dados$folds == k,]
-            
-            categorias.svm.radial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                         kernel = "radial", 
-                                         type="C-classification", epsilon = epsilonsvm, cost = costsvm)     
+
+            categorias.svm.radial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                         kernel = "radial",
+                                         type="C-classification", epsilon = epsilonsvm, cost = costsvm)
             summary(categorias.svm.radial)
-            
-            categorias.svm.radial.pred <- as.character(predict(categorias.svm.radial, 
+
+            categorias.svm.radial.pred <- as.character(predict(categorias.svm.radial,
                                                                newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
             dados[dados$folds == k, 'pred_svm_radial'] <- categorias.svm.radial.pred
         }
-        
+
         cv_svm_radial[counter, 1] <- costsvm
         cv_svm_radial[counter, 2] <- epsilonsvm
         cv_svm_radial[counter, 3] <- mean(dados$class == dados$pred_svm_radial)
@@ -308,13 +308,13 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
-    categorias.svm.radial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                 kernel = "radial", 
-                                 type="C-classification", epsilon = epsilon_opt, cost = cost_opt)     
+
+    categorias.svm.radial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                 kernel = "radial",
+                                 type="C-classification", epsilon = epsilon_opt, cost = cost_opt)
     summary(categorias.svm.radial)
-    
-    categorias.svm.radial.pred <- as.character(predict(categorias.svm.radial, 
+
+    categorias.svm.radial.pred <- as.character(predict(categorias.svm.radial,
                                                        newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
     dados[dados$folds == k, 'pred_svm_radial'] <- categorias.svm.radial.pred
 }
@@ -340,17 +340,17 @@ for (costsvm in list_costs)
         {
             dadosTrain <- dados[dados$folds != k,]
             dadosTest <- dados[dados$folds == k,]
-            
-            categorias.svm.polynomial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                             kernel = "polynomial", 
-                                             type="C-classification", epsilon = epsilonsvm, cost = costsvm)     
+
+            categorias.svm.polynomial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                             kernel = "polynomial",
+                                             type="C-classification", epsilon = epsilonsvm, cost = costsvm)
             summary(categorias.svm.polynomial)
-            
-            categorias.svm.polynomial.pred <- as.character(predict(categorias.svm.polynomial, 
+
+            categorias.svm.polynomial.pred <- as.character(predict(categorias.svm.polynomial,
                                                                    newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
             dados[dados$folds == k, 'pred_svm_polynomial'] <- categorias.svm.polynomial.pred
         }
-        
+
         cv_svm_polynomial[counter, 1] <- costsvm
         cv_svm_polynomial[counter, 2] <- epsilonsvm
         cv_svm_polynomial[counter, 3] <- mean(dados$class == dados$pred_svm_polynomial)
@@ -367,13 +367,13 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
-    categorias.svm.polynomial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                     kernel = "polynomial", 
-                                     type="C-classification", epsilon = epsilon_opt, cost = cost_opt)     
+
+    categorias.svm.polynomial <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                     kernel = "polynomial",
+                                     type="C-classification", epsilon = epsilon_opt, cost = cost_opt)
     summary(categorias.svm.polynomial)
-    
-    categorias.svm.polynomial.pred <- as.character(predict(categorias.svm.polynomial, 
+
+    categorias.svm.polynomial.pred <- as.character(predict(categorias.svm.polynomial,
                                                            newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
     dados[dados$folds == k, 'pred_svm_polynomial'] <- categorias.svm.polynomial.pred
 }
@@ -399,17 +399,17 @@ for (costsvm in list_costs)
         {
             dadosTrain <- dados[dados$folds != k,]
             dadosTest <- dados[dados$folds == k,]
-            
-            categorias.svm.sigmoid <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                          kernel = "sigmoid", 
-                                          type="C-classification", epsilon = epsilonsvm, cost = costsvm)     
+
+            categorias.svm.sigmoid <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                          kernel = "sigmoid",
+                                          type="C-classification", epsilon = epsilonsvm, cost = costsvm)
             summary(categorias.svm.sigmoid)
-            
-            categorias.svm.sigmoid.pred <- as.character(predict(categorias.svm.sigmoid, 
+
+            categorias.svm.sigmoid.pred <- as.character(predict(categorias.svm.sigmoid,
                                                                 newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
             dados[dados$folds == k, 'pred_svm_sigmoid'] <- categorias.svm.sigmoid.pred
         }
-        
+
         cv_svm_sigmoid[counter, 1] <- costsvm
         cv_svm_sigmoid[counter, 2] <- epsilonsvm
         cv_svm_sigmoid[counter, 3] <- mean(dados$class == dados$pred_svm_sigmoid)
@@ -426,13 +426,13 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
-    categorias.svm.sigmoid <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")], 
-                                  kernel = "sigmoid", 
-                                  type="C-classification", epsilon = epsilon_opt, cost = cost_opt)     
+
+    categorias.svm.sigmoid <- svm(formula1, data=dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")],
+                                  kernel = "sigmoid",
+                                  type="C-classification", epsilon = epsilon_opt, cost = cost_opt)
     summary(categorias.svm.sigmoid)
-    
-    categorias.svm.sigmoid.pred <- as.character(predict(categorias.svm.sigmoid, 
+
+    categorias.svm.sigmoid.pred <- as.character(predict(categorias.svm.sigmoid,
                                                         newdata = dadosTest[,colnames(dadosTest) %in% c(list_refs, "class")]))
     dados[dados$folds == k, 'pred_svm_sigmoid'] <- categorias.svm.sigmoid.pred
 }
@@ -444,7 +444,7 @@ mean(dados$class == dados$pred_svm_sigmoid)
 #--- ridge multinomial regression                        ---#
 #-----------------------------------------------------------#
 
-yDados <- data.matrix(dados[,length(list_refs)+4]); head(yDados)
+yDados <- data.matrix(dados[,"class"]); head(yDados)
 xDados <- log(data.matrix(dados[,c(2:(length(list_refs)+1))])); head(xDados)
 
 set.seed(2104)
@@ -457,13 +457,13 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
-    yTrain <- data.matrix(dadosTrain[,length(list_refs)+4])
-    yTest <-  data.matrix(dadosTest[,length(list_refs)+4])
-    
+
+    yTrain <- data.matrix(dadosTrain[,"class"])
+    yTest <-  data.matrix(dadosTest[,"class"])
+
     xTrain <- log(data.matrix(dadosTrain[,c(2:(length(list_refs)+1))]))
     xTest <-   log(data.matrix(dadosTest[,c(2:(length(list_refs)+1))]))
-    
+
     categorias.ridge.pred <- predict(categorias.ridge, s=categorias.ridge$lambda.min, newx=xTest, type='class')
     dados[dados$folds == k, 'pred_ridge'] <- categorias.ridge.pred
 }
@@ -490,13 +490,13 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
+
     yTrain <- data.matrix(dadosTrain[,length(list_refs)+4])
     yTest <-  data.matrix(dadosTest[,length(list_refs)+4])
-    
+
     xTrain <- log(data.matrix(dadosTrain[,c(2:(length(list_refs)+1))]))
     xTest <-   log(data.matrix(dadosTest[,c(2:(length(list_refs)+1))]))
-    
+
     categorias.lasso.pred <- predict(categorias.lasso, s=categorias.lasso$lambda.min, newx=xTest, type='class')
     dados[dados$folds == k, 'pred_lasso'] <- categorias.lasso.pred
 }
@@ -521,11 +521,11 @@ for (alphaelnet in list_alpha)
     categorias.elnet <- cv.glmnet(y = factor(yDados), x = xDados, family="multinomial", alpha=alphaelnet, k = kfolds)
     summary(categorias.elnet)
     elnet_lambda_opt <- categorias.elnet$lambda.min; elnet_lambda_opt
-    
+
     cv_elnet_alpha[contador, 1] <- alphaelnet
     cv_elnet_alpha[contador, 2] <- elnet_lambda_opt
     cv_elnet_alpha[contador, 3] <- min(categorias.elnet$cvm)
-    
+
     contador <- contador + 1
 }
 
@@ -542,13 +542,13 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
+
     yTrain <- data.matrix(dadosTrain[,length(list_refs)+4])
     yTest <-  data.matrix(dadosTest[,length(list_refs)+4])
-    
+
     xTrain <- log(data.matrix(dadosTrain[,c(2:(length(list_refs)+1))]))
     xTest <-   log(data.matrix(dadosTest[,c(2:(length(list_refs)+1))]))
-    
+
     categorias.elnet.pred <- predict(categorias.elnet, s=categorias.elnet$lambda.min, newx=xTest, type='class')
     dados[dados$folds == k, 'pred_elnet'] <- categorias.elnet.pred
 }
@@ -574,21 +574,21 @@ for (ntreesrf in list_ntrees)
         {
             dadosTrain <- dados[dados$folds != k,]
             dadosTest <- dados[dados$folds == k,]
-            
+
             yTrain <- data.matrix(dadosTrain[,length(list_refs)+4])
             yTest <-  data.matrix(dadosTest[,length(list_refs)+4])
-            
+
             xTrain <- log(data.matrix(dadosTrain[,c(2:(length(list_refs)+1))]))
             xTest <-   log(data.matrix(dadosTest[,c(2:(length(list_refs)+1))]))
-            
-            categorias.rfore <- randomForest(y = factor(yTrain), x = xTrain, data=NULL, 
-                                             ntree=ntreesrf, nodesize = nodesizerf, norm.votes=FALSE)    
+
+            categorias.rfore <- randomForest(y = factor(yTrain), x = xTrain, data=NULL,
+                                             ntree=ntreesrf, nodesize = nodesizerf, norm.votes=FALSE)
             summary(categorias.rfore)
-            
+
             categorias.rfore.pred <- as.character(predict(categorias.rfore, newdata = xTest, type = 'response'))
             dados[dados$folds == k, 'pred_rfore'] <- categorias.rfore.pred
         }
-        
+
         cv_rfore[counter, 1] <- ntreesrf
         cv_rfore[counter, 2] <- nodesizerf
         cv_rfore[counter, 3] <- mean(dados$class == dados$pred_rfore)
@@ -605,17 +605,17 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
+
     yTrain <- data.matrix(dadosTrain[,length(list_refs)+4])
     yTest <-  data.matrix(dadosTest[,length(list_refs)+4])
-    
+
     xTrain <- log(data.matrix(dadosTrain[,c(2:(length(list_refs)+1))]))
     xTest <-   log(data.matrix(dadosTest[,c(2:(length(list_refs)+1))]))
-    
-    categorias.rfore <- randomForest(y = factor(yTrain), x = xTrain, data=NULL, 
-                                     ntree=ntrees_opt, nodesize = nodesize_opt, norm.votes=FALSE)      
+
+    categorias.rfore <- randomForest(y = factor(yTrain), x = xTrain, data=NULL,
+                                     ntree=ntrees_opt, nodesize = nodesize_opt, norm.votes=FALSE)
     summary(categorias.rfore)
-    
+
     categorias.rfore.pred <- as.character(predict(categorias.rfore, newdata = xTest, type = 'response'))
     dados[dados$folds == k, 'pred_rfore'] <- categorias.rfore.pred
 }
@@ -644,20 +644,20 @@ for (ntreesgbm in list_ntrees)
         {
             dadosTrain <- dados[dados$folds != k,]
             dadosTest <- dados[dados$folds == k,]
-            
+
             h2o_dadosTrain <- as.h2o(dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")])
             h2o_dadosTrain$class <- as.factor(h2o_dadosTrain$class)
             h2o_dadosTest <- as.h2o(dadosTest[,colnames(dadosTest) %in% list_refs])
-            
-            categorias.h2ogbm <- h2o.gbm(y = "class", x = list_refs, 
+
+            categorias.h2ogbm <- h2o.gbm(y = "class", x = list_refs,
                                          training_frame = h2o_dadosTrain, distribution = "multinomial",
                                          ntrees = ntreesgbm, max_depth = depthgbm);
-            
+
             categorias.h2ogbm.pred <- h2o.predict(categorias.h2ogbm, newdata = h2o_dadosTest)$predict;
-            
+
             dados[dados$folds == k, 'pred_h2ogbm'] <- as.character(as.vector(categorias.h2ogbm.pred));
         }
-        
+
         cv_h2ogbm[counter, 1] <- ntreesgbm
         cv_h2ogbm[counter, 2] <- depthgbm
         cv_h2ogbm[counter, 3] <- mean(dados$class == dados$pred_h2ogbm)
@@ -676,17 +676,17 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
+
     h2o_dadosTrain <- as.h2o(dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")])
     h2o_dadosTrain$class <- as.factor(h2o_dadosTrain$class)
     h2o_dadosTest <- as.h2o(dadosTest[,colnames(dadosTest) %in% list_refs])
-    
-    categorias.h2ogbm <- h2o.gbm(y = "class", x = list_refs, 
+
+    categorias.h2ogbm <- h2o.gbm(y = "class", x = list_refs,
                                  training_frame = h2o_dadosTrain, distribution = "multinomial",
                                  ntrees = ntrees_opt, max_depth = depth_opt);
-    
+
     categorias.h2ogbm.pred <- h2o.predict(categorias.h2ogbm, newdata = h2o_dadosTest)$predict;
-    
+
     dados[dados$folds == k, 'pred_h2ogbm'] <- as.character(as.vector(categorias.h2ogbm.pred));
 }
 
@@ -716,20 +716,20 @@ for (ntreesrfore in list_ntrees)
         {
             dadosTrain <- dados[dados$folds != k,]
             dadosTest <- dados[dados$folds == k,]
-            
+
             h2o_dadosTrain <- as.h2o(dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")])
             h2o_dadosTrain$class <- as.factor(h2o_dadosTrain$class)
             h2o_dadosTest <- as.h2o(dadosTest[,colnames(dadosTest) %in% list_refs])
-            
-            categorias.h20rfore <- h2o.randomForest(y = "class", x = list_refs, 
-                                                    training_frame = h2o_dadosTrain, 
-                                                    ntrees = ntreesrfore, max_depth = depthrfore)      
-            
+
+            categorias.h20rfore <- h2o.randomForest(y = "class", x = list_refs,
+                                                    training_frame = h2o_dadosTrain,
+                                                    ntrees = ntreesrfore, max_depth = depthrfore)
+
             categorias.h20rfore.pred <- h2o.predict(categorias.h20rfore, newdata = h2o_dadosTest)$predict;
-            
+
             dados[dados$folds == k, 'pred_h2orfore'] <- as.character(as.vector(categorias.h20rfore.pred));
         }
-        
+
         cv_h2orfore[counter, 1] <- ntreesrfore
         cv_h2orfore[counter, 2] <- depthrfore
         cv_h2orfore[counter, 3] <- mean(dados$class == dados$pred_h2orfore)
@@ -748,17 +748,17 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
+
     h2o_dadosTrain <- as.h2o(dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")])
     h2o_dadosTrain$class <- as.factor(h2o_dadosTrain$class)
     h2o_dadosTest <- as.h2o(dadosTest[,colnames(dadosTest) %in% list_refs])
-    
-    categorias.h20rfore <- h2o.randomForest(y = "class", x = list_refs, 
-                                            training_frame = h2o_dadosTrain, 
-                                            ntrees = ntrees_opt, max_depth = depth_opt)      
-    
+
+    categorias.h20rfore <- h2o.randomForest(y = "class", x = list_refs,
+                                            training_frame = h2o_dadosTrain,
+                                            ntrees = ntrees_opt, max_depth = depth_opt)
+
     categorias.h20rfore.pred <- h2o.predict(categorias.h20rfore, newdata = h2o_dadosTest)$predict;
-    
+
     dados[dados$folds == k, 'pred_h2orfore'] <- as.character(as.vector(categorias.h20rfore.pred));
 }
 
@@ -778,16 +778,16 @@ for (k in 1:kfolds)
 {
     dadosTrain <- dados[dados$folds != k,]
     dadosTest <- dados[dados$folds == k,]
-    
+
     h2o_dadosTrain <- as.h2o(dadosTrain[,colnames(dadosTrain) %in% c(list_refs, "class")])
     h2o_dadosTrain$class <- as.factor(h2o_dadosTrain$class)
     h2o_dadosTest <- as.h2o(dadosTest[,colnames(dadosTest) %in% list_refs])
-    
-    categorias.h2odpl <- h2o.deeplearning(y = "class", x = list_refs, 
+
+    categorias.h2odpl <- h2o.deeplearning(y = "class", x = list_refs,
                                           training_frame = h2o_dadosTrain, distribution = "multinomial");
-    
+
     categorias.h2odpl.pred <- h2o.predict(categorias.h2odpl, newdata = h2o_dadosTest)$predict;
-    
+
     dados[dados$folds == k, 'pred_h2odpl'] <- as.character(as.vector(categorias.h2odpl.pred));
 }