Add 'require(pkgname)' in examples and fix some typos

R/ConnectionWeights.R

@@ -1,7 +1,7 @@
-#' Connection Weights method
+#' Connection weights method
 #'
 #' @description
-#' This class implements the *Connection Weights* method investigated by
+#' This class implements the *Connection weights* method investigated by
 #' Olden et al. (2004), which results in a relevance score for each input
 #' variable. The basic idea is to multiply all path weights for each
 #' possible connection between an input feature and the output node and then
@@ -14,7 +14,7 @@
 #'
 #' In this package, we extended this method to a local method inspired by the
 #' method *Gradient\eqn{\times}Input* (see [`Gradient`]). Hence, the local variant is
-#' simply the point-wise product of the global *Connection Weights* method and
+#' simply the point-wise product of the global *Connection weights* method and
 #' the input data. You can use this variant by setting the `times_input`
 #' argument to `TRUE` and providing input data.
 #'
@@ -39,7 +39,7 @@ ConnectionWeights <- R6Class(
   public = list(
     #' @field times_input (`logical(1)`)\cr
     #' This logical value indicates whether the results from
-    #' the *Connection Weights* method were multiplied by the provided input
+    #' the *Connection weights* method were multiplied by the provided input
     #' data or not. Thus, this value specifies whether the original global
     #' variant of the method or the local one was applied. If the value is
     #' `TRUE`, then data is provided in the field `data`.
@@ -52,7 +52,7 @@ ConnectionWeights <- R6Class(
     #'
     #' @param times_input (`logical(1)`)\cr
     #' Multiplies the results with the input features.
-    #' This variant turns the global *Connection Weights* method into a local
+    #' This variant turns the global *Connection weights* method into a local
     #' one. Default: `FALSE`.\cr
     initialize = function(converter,
                           data = NULL,

R/DeepLift.R

@@ -1,7 +1,7 @@
-#' @title Deep Learning Important Features (DeepLift)
+#' @title Deep learning important features (DeepLift)
 #'
 #' @description
-#' This is an implementation of the \emph{Deep Learning Important Features
+#' This is an implementation of the \emph{deep learning important features
 #' (DeepLift)} algorithm introduced by Shrikumar et al. (2017). It's a local
 #' method for interpreting a single element \eqn{x} of the dataset concerning
 #' a reference value \eqn{x'} and returns the contribution of each input
@@ -10,7 +10,7 @@
 #' decompose the difference-from-reference prediction with respect to the
 #' input features, i.e.,
 #' \deqn{\Delta y = y - y'  = \sum_i C(x_i).}
-#' Compared to \emph{Layer-wise Relevance Propagation} (see [LRP]), the
+#' Compared to \emph{Layer-wise relevance propagation} (see [LRP]), the
 #' DeepLift method is an exact decomposition and not an approximation, so we
 #' get real contributions of the input features to the
 #' difference-from-reference prediction. There are two ways to handle

R/LRP.R

@@ -1,7 +1,7 @@
-#' @title Layer-wise Relevance Propagation (LRP)
+#' @title Layer-wise relevance propagation (LRP)
 #'
 #' @description
-#' This is an implementation of the \emph{Layer-wise Relevance Propagation
+#' This is an implementation of the \emph{layer-wise relevance propagation
 #' (LRP)} algorithm introduced by Bach et al. (2015). It's a local method for
 #' interpreting a single element of the dataset and calculates the relevance
 #' scores for each input feature to the model output. The basic idea of this

R/innsight.R

@@ -13,10 +13,10 @@
 #' This package implements several model-specific interpretability
 #' (feature attribution) methods based on neural networks in R, e.g.,
 #'
-#' * *Layer-wise Relevance Propagation ([LRP])*
+#' * *Layer-wise relevance propagation ([LRP])*
 #'   * Including propagation rules: \eqn{\epsilon}-rule and
 #'   \eqn{\alpha}-\eqn{\beta}-rule
-#' * *Deep Learning Important Features ([DeepLift])*
+#' * *Deep learning important features ([DeepLift])*
 #'   * Including propagation rules for non-linearities: *Rescale* rule and
 #'  *RevealCancel* rule
 #' * Gradient-based methods:

man-roxygen/examples-ConnectionWeights.R

@@ -26,69 +26,72 @@
 #' plot(cw)
 #'
 #' #----------------------- Example 2: Neuralnet ------------------------------
-#' library(neuralnet)
-#' data(iris)
-#'
-#' # Train a Neural Network
-#' nn <- neuralnet((Species == "setosa") ~ Petal.Length + Petal.Width,
-#'   iris,
-#'   linear.output = FALSE,
-#'   hidden = c(3, 2), act.fct = "tanh", rep = 1
-#' )
+#' if (require("neuralnet")) {
+#'   library(neuralnet)
+#'   data(iris)
 #'
-#' # Convert the trained model
-#' converter <- Converter$new(nn)
+#'   # Train a Neural Network
+#'   nn <- neuralnet((Species == "setosa") ~ Petal.Length + Petal.Width,
+#'     iris,
+#'     linear.output = FALSE,
+#'     hidden = c(3, 2), act.fct = "tanh", rep = 1
+#'   )
 #'
-#' # Apply the Connection Weights method
-#' cw <- ConnectionWeights$new(converter)
+#'   # Convert the trained model
+#'   converter <- Converter$new(nn)
 #'
-#' # Get the result as a torch tensor
-#' get_result(cw, type = "torch.tensor")
+#'   # Apply the Connection Weights method
+#'   cw <- ConnectionWeights$new(converter)
 #'
-#' # Plot the result
-#' plot(cw)
+#'   # Get the result as a torch tensor
+#'   get_result(cw, type = "torch.tensor")
 #'
+#'   # Plot the result
+#'   plot(cw)
+#' }
 #' @examplesIf keras::is_keras_available() & torch::torch_is_installed()
 #' # ------------------------- Example 3: Keras -------------------------------
-#' library(keras)
-#'
-#' # Make sure keras is installed properly
-#' is_keras_available()
-#'
-#' data <- array(rnorm(10 * 32 * 32 * 3), dim = c(10, 32, 32, 3))
-#'
-#' model <- keras_model_sequential()
-#' model %>%
-#'   layer_conv_2d(
-#'     input_shape = c(32, 32, 3), kernel_size = 8, filters = 8,
-#'     activation = "softplus", padding = "valid") %>%
-#'   layer_conv_2d(
-#'     kernel_size = 8, filters = 4, activation = "tanh",
-#'     padding = "same") %>%
-#'   layer_conv_2d(
-#'     kernel_size = 4, filters = 2, activation = "relu",
-#'     padding = "valid") %>%
-#'   layer_flatten() %>%
-#'   layer_dense(units = 64, activation = "relu") %>%
-#'   layer_dense(units = 16, activation = "relu") %>%
-#'   layer_dense(units = 2, activation = "softmax")
-#'
-#' # Convert the model
-#' converter <- Converter$new(model)
-#'
-#' # Apply the Connection Weights method
-#' cw <- ConnectionWeights$new(converter)
-#'
-#' # Get the head of the result as a data.frame
-#' head(get_result(cw, type = "data.frame"), 5)
-#'
-#' # Plot the result for all classes
-#' plot(cw, output_idx = 1:2)
-#'
+#' if (require("keras")) {
+#'   library(keras)
+#'
+#'   # Make sure keras is installed properly
+#'   is_keras_available()
+#'
+#'   data <- array(rnorm(10 * 32 * 32 * 3), dim = c(10, 32, 32, 3))
+#'
+#'   model <- keras_model_sequential()
+#'   model %>%
+#'     layer_conv_2d(
+#'       input_shape = c(32, 32, 3), kernel_size = 8, filters = 8,
+#'       activation = "softplus", padding = "valid") %>%
+#'     layer_conv_2d(
+#'       kernel_size = 8, filters = 4, activation = "tanh",
+#'       padding = "same") %>%
+#'     layer_conv_2d(
+#'       kernel_size = 4, filters = 2, activation = "relu",
+#'       padding = "valid") %>%
+#'     layer_flatten() %>%
+#'     layer_dense(units = 64, activation = "relu") %>%
+#'     layer_dense(units = 16, activation = "relu") %>%
+#'     layer_dense(units = 2, activation = "softmax")
+#'
+#'   # Convert the model
+#'   converter <- Converter$new(model)
+#'
+#'   # Apply the Connection Weights method
+#'   cw <- ConnectionWeights$new(converter)
+#'
+#'   # Get the head of the result as a data.frame
+#'   head(get_result(cw, type = "data.frame"), 5)
+#'
+#'   # Plot the result for all classes
+#'   plot(cw, output_idx = 1:2)
+#' }
 #' @examplesIf torch::torch_is_installed() & Sys.getenv("RENDER_PLOTLY", unset = 0) == 1
 #' #------------------------- Plotly plots ------------------------------------
-#'
-#' # You can also create an interactive plot with plotly.
-#' # This is a suggested package, so make sure that it is installed
-#' library(plotly)
-#' plot(cw, as_plotly = TRUE)
+#' if (require("plotly")) {
+#'   # You can also create an interactive plot with plotly.
+#'   # This is a suggested package, so make sure that it is installed
+#'   library(plotly)
+#'   plot(cw, as_plotly = TRUE)
+#' }

man-roxygen/examples-Converter.R

@@ -21,49 +21,53 @@
 #'
 #'
 #' #----------------------- Example 2: Neuralnet ------------------------------
-#' library(neuralnet)
-#' data(iris)
-#'
-#' # Train a neural network
-#' nn <- neuralnet((Species == "setosa") ~ Petal.Length + Petal.Width,
-#'   iris,
-#'   linear.output = FALSE,
-#'   hidden = c(3, 2), act.fct = "tanh", rep = 1
-#' )
+#' if (require("neuralnet")) {
+#'   library(neuralnet)
+#'   data(iris)
+#'
+#'   # Train a neural network
+#'   nn <- neuralnet((Species == "setosa") ~ Petal.Length + Petal.Width,
+#'     iris,
+#'     linear.output = FALSE,
+#'     hidden = c(3, 2), act.fct = "tanh", rep = 1
+#'   )
 #'
-#' # Convert the model
-#' converter <- Converter$new(nn)
+#'   # Convert the model
+#'   converter <- Converter$new(nn)
 #'
-#' # Print all the layers
-#' converter$model$modules_list
+#'   # Print all the layers
+#'   converter$model$modules_list
+#' }
 #'
 #' @examplesIf keras::is_keras_available() & torch::torch_is_installed()
 #' #----------------------- Example 3: Keras ----------------------------------
-#' library(keras)
-#'
-#' # Make sure keras is installed properly
-#' is_keras_available()
-#'
-#' # Define a keras model
-#' model <- keras_model_sequential() %>%
-#'   layer_conv_2d(
-#'     input_shape = c(32, 32, 3), kernel_size = 8, filters = 8,
-#'     activation = "relu", padding = "same") %>%
-#'   layer_conv_2d(
-#'     kernel_size = 8, filters = 4,
-#'     activation = "tanh", padding = "same") %>%
-#'   layer_conv_2d(
-#'     kernel_size = 4, filters = 2,
-#'     activation = "relu", padding = "same") %>%
-#'   layer_flatten() %>%
-#'   layer_dense(units = 64, activation = "relu") %>%
-#'   layer_dense(units = 1, activation = "sigmoid")
-#'
-#' # Convert this model and save model as list
-#' converter <- Converter$new(model, save_model_as_list = TRUE)
-#'
-#' # Print the converted model as a named list
-#' str(converter$model_as_list, max.level = 1)
+#' if (require("keras")) {
+#'   library(keras)
+#'
+#'   # Make sure keras is installed properly
+#'   is_keras_available()
+#'
+#'   # Define a keras model
+#'   model <- keras_model_sequential() %>%
+#'     layer_conv_2d(
+#'       input_shape = c(32, 32, 3), kernel_size = 8, filters = 8,
+#'       activation = "relu", padding = "same") %>%
+#'     layer_conv_2d(
+#'       kernel_size = 8, filters = 4,
+#'       activation = "tanh", padding = "same") %>%
+#'     layer_conv_2d(
+#'       kernel_size = 4, filters = 2,
+#'       activation = "relu", padding = "same") %>%
+#'     layer_flatten() %>%
+#'     layer_dense(units = 64, activation = "relu") %>%
+#'     layer_dense(units = 1, activation = "sigmoid")
+#'
+#'   # Convert this model and save model as list
+#'   converter <- Converter$new(model, save_model_as_list = TRUE)
+#'
+#'   # Print the converted model as a named list
+#'   str(converter$model_as_list, max.level = 1)
+#' }
 #'
 #' @examplesIf torch::torch_is_installed()
 #' #----------------------- Example 4: List  ----------------------------------