R/help.R

In rattle: Graphical User Interface for Data Science in R

# Gnome R Data Miner: GNOME interface to R for Data Mining
#
# Time-stamp: <2018-07-01 20:24:22 Graham.Williams@togaware.com>
#
# Help Menu
#
# Copyright (c) 2009 Togaware Pty Ltd
#
# This files is part of Rattle.
#
# Rattle is free software: you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 2 of the License, or
# (at your option) any later version.
#
# Rattle is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
# General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Rattle. If not, see <https://www.gnu.org/licenses/>.

popupTextviewHelpWindow <- function(topic, package)
{
  # We can specify the package so as to be specific about which
  # function we want help on.
  
  pkg <- ifelse(missing(package), "", sprintf("package=%s, ", package))

  # The help_type option was introduced in 2.10.0
  
  html <- ifelse(getRversion()<="2.10.0", "htmlhelp=TRUE", "help_type='html'")

  collectOutput(sprintf("help(%s, %s%s)", topic, pkg, html), TRUE)
}

showHelpPlus <- function(..., extra=Rtxt("Would you like to view the R help?"))
{
  if (! questionDialog(paste(gsub(" <<>> ", "\n\n",
                                  gsub("\n", " ", ...)),
                             extra, sep="\n\n")))
    return(FALSE)
  else
    return(TRUE)
}

showHelp <- function(msg)
{
  infoDialog(paste(gsub(" <<>> ", "\n\n", gsub("\n", " ", msg))))
}

viewDocMsg <- function(doc)
{
  return(sprintf(Rtxt("view documentation for '%s'"), doc))
}

on_help_general_activate <- function(action, window)
{
  showHelp(Rtxt("Rattle is a graphical user interface for data mining",
                "written in GNOME and R. R is an environment for statistical computing.",
                "They are all free software licensed under the GNU General",
                "Public License (GPL).",
                "<<>>",
                "Interaction with Rattle logically proceeds by progressing through the Tabs:",
                "first load in some Data and select variables for exploring and mining,",
                "and possibly sampling the data, Explore the data, apply statistical Tests,",
                "build your Models,",
                "and Evaluate them. For any tab, the modus operandi is to configure",
                "the options available and then click the Execute button (or F2) to perform",
                "the appropriate tasks. Note that the tasks are NOT performed until",
                "the Execute button (or F2 or the Execute menu item under Tools) is clicked.",
                "<<>>",
                "The Status Bar at the bottom of the window indicates when the action",
                "is completed. Messages from R (e.g., error messages - although Rattle does",
                "attempt to catch them first) will appear in the R console",
                "from where you started Rattle. The corresponding R Code will",
                "appear in the Log tab.",
                "This allows you to review the R commands",
                "that perform the corresponding data mining tasks. Even better, you can save the",
                "code from the Log tab as an R script file and copmletely replicate your interactions",
                "with Rattle, sipmly by running the script. You can also paste the code into the same R Console",
                "from which Rattle is running, and execute the commands directly, then continue back in Rattle.",
                "This allows you to use Rattle to do the basics, and then where you",
                "need more sophistication, go into R directly. Rattle uses a variable called",
                "crs to store its current state, and you can modify this directly.",
                "<<>>",
                "Rattle works well for",
                "binary classification problems (with 0/1 or a two level variable",
                "as the outcomes for the Target variable) with support for",
                "mulitnomial classification and regression tasks.",
                "<<>>",
                "The most we guarantee",
                "is that there are bugs! When you find one, or a misfeature or",
                "something else you would like Rattle to do, please do email",
                "support@togaware.com.",
                "<<>>",
                "Enjoy."))
}

on_help_project_menuitem_activate <- function(action, window)
{
  showHelp(Rtxt("A Project is where all the related information from Rattle is",
                "stored in a file, so that at a later time we can resume our work",
                "in Rattle.",
                "<<>>",
                "Projects can be saved and loaded. A Rattle project has the",
                "filename extension of .rattle"))
}

########################################################################
# DATA

on_help_nomenclature_data_activate <- function(action, window)
{
  showHelp(Rtxt("There are many",
           "different nomenclatures being used in data mining, deriving from the many",
           "different contributory fields. Here, we attempt to stay with a single,",
           "consistent nomenclature.",
           "<<>>",
           "A dataset consists of observations described using variables,",
           "which might consist of a mixture of input variables and output variables,",
           "either of which may be categoric or numeric.",
           "<<>>",
           "dataset = A collection of data.",
           "<<>>",
           "observation = An object or entity of interest, described by variables.",
           "Also called a record, object, row or entity.",
           "<<>>",
           "variable = The data items used to describe an entity.",
           "Also called an attribute, feature or column.",
           "<<>>",
           "input variable = A measured or preset data item.",
           "Also called predictor, independent variable, observed variable,",
           "or descriptive variable.",
           "<<>>",
           "output variable = A variable possibly influenced by the input variables.",
           "Also called response or dependent variable.",
           "<<>>",
           "categoric variable = A variable that takes on a value from a fixed",
           "set of values. In R these are called factors and the possible values",
           "are referred to as the levels of the factor.",
           "<<>>",
           "numeric variable = A variable that has values that are integers or real",
           "numbers."))
}

on_help_data_fex_activate <- function(action, window)
{
  showHelp(Rtxt("The FEX option of the Data tab provides a direct linkage to Web Focus",
                "as the source of our dataset for analysis."))
}

on_help_csv_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Data can be loaded from",
                         "a comma separated value CSV file, as might be generated",
                         "by spreadsheets and databases,",
                         "including Excel, Gnumeric, SAS/EM, QueryMan,",
                         "and many other applications.",
                         "This is a good option for importing data.",
                         "<<>>",
                         "The CSV file is assumed to begin with a header row,",
                         "listing the names of the variables. ",
                         "The remainder of the file is expected to consist of",
                         "rows of data that record",
                         "information about the observations,",
                         "with fields generally separated by commas",
                         "recording the values of the variables for this observation.",
                         "<<>>",
                         "Use the Separator box to choose a separator",
                         "other than the default comma.",
                         "A common alternative is a tab,",
                         "or simply leave it blank to have",
                         "any white space act as a separator.",
                         "<<>>",
                         "A URL can be supplied in the Location:",
                         "text box so that a CSV file can be",
                         "loaded from the network.",
                         "<<>>",
                         "The corresponding R code uses the",
                         "simple 'read.csv' function.")))
    popupTextviewHelpWindow("read.csv") }

on_help_arff_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Data can be loaded from",
                   "an Attribute-Relation File Format, or ARFF, file",
                   "(beginning with version 2.5.0 of R).",
                   "ARFF is an ASCII text file format",
                   "that is essentially a CSV file with a header that describes the",
                   "meta-data. ARFF was developed for use in the Weka machine learning",
                   "software and there are quite a few datasets in this format now.",
                   "<<>>",
                   "The corresponding R code uses the 'read.arff' function from the",
                   "foreign package.")))
  {
    popupTextviewHelpWindow("read.arff", "foreign")
  }
}

on_help_rdata_file_activate <- function(action, window)
{
  showHelp(Rtxt("Choose this if you have data stored in an R dataset",
                "(usually with a filename extension of .Rdata).",
                "The named file will be loaded and any data frames found in there will",
                "be listed for selection."))
}

on_help_data_library_activate <- function(action, window)
{
  showHelp(Rtxt("The Library option of the Data tab provides access to all sample datasets",
           "provided by the various R packages in your installation. When selected, all of the",
           "packages are scanned for the datasets they provide. You can then choose a dataset",
           "from the drop down menu."))
}

on_help_rdataset_activate <- function(action, window)
{
  showHelp(Rtxt("Datasets already loaded into R can be used",
           "(although a copy is taken, with memory implications).",
           "Only data frames are currently supported, and ",
           "the names of all of the available data frames will be listed.",
           "<<>>",
           "The data frames need to be constructed in the same R session",
           "that is running Rattle (i.e., the same R Console in which you",
           "sourced the Rattle package). This provides much more flexibility in",
           "loading data into Rattle, than is provided directly through the actual Rattle",
           "interface. For example, you may want to use the SQLLite package to load",
           "data from a database directly."))
}

on_help_odbc_activate <- function(action, window)
{
  if(showHelpPlus(Rtxt("Rattle can establish a connection to a database",
                  "through the RODBC package. Tables avilable in the database will then be",
                  "listed for selection.")))
  {
    popupTextviewHelpWindow("RODBC", "RODBC")
  }
}

on_help_data_corpus_activate <- function(action, window)
{
  showHelp(Rtxt("The Corpus option of the Data tab will load and process a folder of",
           "documents, referred to as a corpus. Each document will be processed according to the",
           "options specified, to end up with a dataset with, in the simplest case, variables",
           "corresponding to the keywords identified in the documents, and the observations",
           "corresponding to each documents. The full functionality of Rattle can then be deployed",
           "on this dataset.",
           "<<>>",
           "The tm package of R provides the underlying functionality to process a corpus."))
}

on_help_roles_activate <- function(action, window)
{
  showHelp(Rtxt("The Data tab allows you to select roles for the",
           "variables.",
           "<<>>",
           "By default, all variables have an Input role, except for any variables",
           "that have constant value, or categorics with as many values as there",
           "are rows (identifier variables). These will be marked as Ignore.",
           "<<>>",
           "One variable may also be identified as the Target (the first or last",
           "categoric by default).",
           "<<>>",
           "Modify the roles as appropriate for each variable.",
           "<<>>",
           "Input: Used for modeling.",
           "<<>>",
           "Target: Output for modeling.",
           "<<>>",
           "Risk: A variable used in the Risk Chart",
           "<<>>",
           "Ident: An identifier for unique observations in the data set.",
           "<<>>",
           "Ignore: Do not use this variable",
           "<<>>",
           "The Input and Ignore buttons can be used to operate on a selection.",
           "A Shift-Click in the variable list will select all variables from the",
           "last click to current variable. A Ctrl-Click will add the current",
           "variable to those selected."))
}

on_help_weight_calculator_activate <- function(action, window)
{
  showHelp(Rtxt("Weights are used by variable modellers to identify",
           "some observations as more important than others. The Weights Calculator",
           "can be used to specify a formula in terms of the variables in the dataset.",
           "You can list just a variable name, or",
           "any formula can be used, as long as it is a valid R formula - R will be",
           "asked to evaluate the formula.",
           "<<>>",
           "An example might be 'abs(Rev)/max(Rev)*10+1' which takes the absolute",
           "value of a variable called Rev, divides it by the maximum value of Rev in the",
           "dataset, times 10, adding 1 to it to give numbers from 1 up."))
}

on_help_sample_activate <- function(action, window)
{
  showHelp(Rtxt("Partitioning (sampling) is activated by default,",
                "randomly choosing 70% of the",
                "data for a training dataset,",
                "15% for a validation dataset, and",
                "15% for a test dataset. The training dataset",
                "is used to build models, the validation dataset",
                "to tune the models,",
                "and the test dataset is used to evaluate the",
                "models on otherwise unseen data.",
                "<<>>",
                "A new random sample is extracted each time",
                "the tab is executed. However,",
                "you will get the same random sample each time,",
                "for a given seed. Changing the",
                "seed allows different random samples to be extracted.",
                "This could be useful in",
                "testing the sensitivity of modeling with different",
                "training sets."))
}

########################################################################
# EXPLORE TAB

on_help_summary_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("A summary of the dataset includes various pieces of",
                   "information about each of the variables of the dataset.",
                   "<<>>",
                   "For numeric data, this",
                   "can include the minimum, maximum, median (the value of the variable at the",
                   "midpoint of the dataset), mean (the average value of the variable),",
                   "and the first and third quartiles (25 percent of the data has values",
                   "below the first",
                   "quartile, and another 25 percent of the data has values above the third quartile).",
                   "<<>>",
                   "For categoric data the frequency distribution across the values is listed.",
                   "If there are too many possible values, then only the top few are listed, with",
                   "the remainder counted as Other.",
                   "<<>>",
                   "The R function summary() is used for the summary.",
                   "<<>>",
                   "Additional or differently presented summary information is provided",
                   "through additional options. Describe produces a similar summary presented",
                   "differently. For numeric variables, the Basic statistics can be obtained,",
                   "including kurtosis and skewness.",
                   "<<>>",
                   "The kurtosis is a measure of the nature of the peaks",
                   "in the distribution of the data. A high kurtosis indicates a sharper peak",
                   "and fatter tails while a lower kurtosis indicates a more rounded peak",
                   "with wider shoulders.",
                   "<<>>",
                   "The skewness indicates the asymmetry of the distribution. A positive skew",
                   "indicates that the tail to the right is longer, and a negative skew that the",
                   "tail to the left is longer.",
                   "<<>>",
                   "The fBasics package is used for the Basic summary and",
                   "the kurtosis and skewness.")))
    {
      popupTextviewHelpWindow("summary", "base")
      if (packageIsAvailable("Hmisc", viewDocMsg("describe")))
      {
        popupTextviewHelpWindow("describe", "Hmisc")
      }
      if (packageIsAvailable("fBasics", viewDocMsg("basicStats")))
      {
        popupTextviewHelpWindow("basicStats", "fBasics")
      }
    }
}

on_help_distributions_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Choose from various plot types to display",
                        "information about the distributions of data.")))
    popupTextviewHelpWindow("boxplot")
}

## on_help_latticist_activate <- function(action, window)
## {
##   if (showHelpPlus(Rtxt("The Latticist application is used to visually explore",
##                         "a dataset. Latticist is an R-based interactive",
##                         "data visualizer.")))
##     if (packageIsAvailable("latticist", viewDocMsg("latticist")))
##       {
##         popupTextviewHelpWindow("latticist", "latticist")
##       }
## }

on_help_ggobi_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("GGobi application is used to visually explore",
                        "a dataset. GGobi is a very powerful interactive visualizer.",
                        "The separate GGobi application will need to have been",
                        " installed, as well as the rggobi R package.")))
    if (packageIsAvailable("rggobi", viewDocMsg("rggobi")))
      {
        popupTextviewHelpWindow("rggobi", "rggobi")
      }
}

on_help_ggraptr_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("GGRaptr implements a shiny-based interactive tool to visually explore",
                        "a dataset using GGPlot2.")))
    if (packageIsAvailable("ggraptR", viewDocMsg("ggraptR")))
      {
        popupTextviewHelpWindow("ggraptR", "ggraptR")
      }
}

on_help_correlation_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("A pair wise correlation between each numeric variable",
                        "is calculated and displayed numerically in the text window",
                        "whilst a graphic plot is also generated. The plot uses circles",
                        "and colour to indicate the strength of any correlation.",
                        "<<>>",
                        "The R function cor() is used to produce the",
                        "correlation data.")))
    popupTextviewHelpWindow("cor")
}

on_help_hierarchical_correlation_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("A hierarchical cluster of the correlations between",
                        "the variables of the dataset is generated, and",
                        "presented pictorially as a dendrogram.  From the",
                        "dendrogram you can",
                        "see groups of variables that are highly correlated.",
                        "The code uses the",
                        "cor() function to generate the correlations between",
                        "the variables, the",
                        "hclust() function to perform the hierarchical",
                        "clustering, and converts",
                        "the result to a dendrogram, using as.dendrogram(),",
                        "for plotting.")))
  {
    popupTextviewHelpWindow("cor")
    popupTextviewHelpWindow("hclust")
    popupTextviewHelpWindow("dendrogram")
  }
 
}

on_help_principal_components_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Principal components analysis identifies",
                        "a collection of derived variables (expressed as a linear combination",
                        "of the other variables) that account for the variance",
                        "in the data. Often, the first few components account for the majority",
                        "of the variation.",
                        "<<>>",
                        "After performing the analysis two plots will appear. The bar chart or scree plot",
                        "shows the importance of each component. The importance is based on how much",
                        "variance each component explains. Generally, we can use this plot to how many",
                        "principle components we may want to keep if we were to use them in modeling.",
                        "<<>>",
                        "The second plot (a biplot) remaps the data points from their original",
                        "coordinates to coordinates of the first two principal coordinates.",
                        "The vectors drawn give an indication of how much",
                        "of a role each variable plays in each of the two components, showing their correlation",
                        "to the components. The axes are labeled with the correlation, to be",
                        "interpreted for the variables, and the values of the principal",
                        "components, to be interpreted for the data points.",
                        "<<>>",
                        "There will be as many components as there are (numeric) variables in",
                        "the dataset, but by discarding those components contributing very",
                        "little, you may end up with fewer variables for modeling. The textual information",
                        "provided can be used to guide the choice of which variables are included.",
                        "The final table will clearly identify how much of the variation in the data",
                        "is accounted for by each component.",
                        "<<>>",
                        "Interpretability may reduce through using the derived variables rather",
                        "than the original variables, so you may like to instead identify those",
                        "variables that contribute most to the first few principal components.",
                        "<<>>",
                        "The prcomp() function is used to generate the principal components",
                        "which are then displayed in the text view and the relative importance",
                        "of the components is plotted.",
                        "<<>>",
                        "Note that only numeric data is included in the analysis.")))
    popupTextviewHelpWindow("prcomp")
}

########################################################################
# TEST TAB

on_help_test_kolmogorov_smirnov_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The Kolmogorov-Smirnov test is a test to determine whether",
                   "two samples are similarly distributed, without saying what kind of distribution",
                   "they have.",
                   "<<>>",
                   "The fBasics package provides the ks2Test function to perform the",
                   "Kolmogorov-Smirnov Test.")))
  {
    if (packageIsAvailable("fBasics", viewDocMsg("k2test")))
    {
      popupTextviewHelpWindow("ks2Test", "fBasics")
    }
  }
}

on_help_test_wilcoxon_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The Wilcoxon test, also known as the Mann-Whitney test,",
                   "is analogous to the two-sample t-test, but performed on the rankings of the",
                   "combined data sets instead of on the actual measure. If the observations",
                   "rankings are not different, then the samples are not different. Because it is",
                   "performed on the rankings, it is more sensitive about the location of the",
                   "distribution, i.e. to the median (not the mean as in the T-Test).",
                   "<<>>",
                   "The fBasics package provides the wilcox.test function to perform the",
                   "Wilcoxon Test.")))
  {
    if (packageIsAvailable("fBasics", viewDocMsg("wilcox.test")))
    {
      popupTextviewHelpWindow("wilcox.test", "fBasics")
    }
  }
}

on_help_test_t_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The T-test is the most commonly used test to determine whether",
                   "the means of two normally distributed samples are of equal sizes. The mean is a",
                   "measure of the location of the distribution. If the two populations are normal",
                   "(bell shaped) and their means are different, then the two bell shapes will be",
                   "offset from one another, indicating that the two samples are different. If the",
                   "means are equal the bell shapes will overlap.",
                   "<<>>",
                   "The fBasics package provides the locationTest function with the 't' method",
                   "to perform the T-test.")))
  {
    if (packageIsAvailable("fBasics", viewDocMsg("locationTest")))
    {
      popupTextviewHelpWindow("locationTest", "fBasics")
    }
  }
}
on_help_test_f_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The F-test is used to test whether the variance of the data",
                   "from two normally distributed samples is the same.",
                   "<<>>",
                   "The fBasics package provides the varianceTest function with the 'varf' method",
                   "to perform the F-test.")))
  {
    if (packageIsAvailable("fBasics", viewDocMsg("varianceTest")))
    {
      popupTextviewHelpWindow("varianceTest", "fBasics")
    }
  }
}

on_help_test_correlation_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The Correlation test is used to test for the existence of",
                   "a linear relationship between the two variables. Only Pearson's Correlation Test",
                   "is performed in the Test tab.",
                   "<<>>",
                   "The fBasics package provides the correlationTest function",
                   "to perform the correlation test.")))
  {
    if (packageIsAvailable("fBasics", viewDocMsg("correlationTest")))
    {
      popupTextviewHelpWindow("correlationTest", "fBasics")
    }
  }
}

on_help_test_wilcoxon_signed_rank_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The Wilcoxon signed rank test is used to test two related",
                   "samples, such as matched pairs or before and after tests, to determine if the",
                   "repeated measurements on the same individuals are the same.",
                   "<<>>",
                   "The fBasics package provides the wilcox.test function with the paired option",
                   "to perform the Wilcoxon signed rank test.")))
  {
    if (packageIsAvailable("fBasics", viewDocMsg("wilcox.test")))
    {
      popupTextviewHelpWindow("wilcox.test", "fBasics")
    }
  }
}

########################################################################
# TRANSFORM TAB

on_help_normalise_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Rescaling options transforms a variable by recoding its",
                   "values to another set of values, such as a set that has a mean of 0 and",
                   "standard deviation of 1. Often we do this so that all of our variables",
                   "have a very similar spread, and perhaps distribution. This can then avoid",
                   "biases in various algorithms, such as in clustering where a distance measure",
                   "is often used.",
                   "<<>>",
                   "Various rescalings are supported, with the rescaler function from the reshape",
                   "package used in a number of cases.",
                   "<<>>",
                   "The By Group Transform segments and recodes a numeric variable to the 0-100 range.")))
  {
    if (packageIsAvailable("reshape", viewDocMsg("rescaler")))
    {
      popupTextviewHelpWindow("rescaler", "reshape")
    }
  }
}

on_help_transform_recenter_activate <- function(action, window)
{
  showHelp(Rtxt("Recenter performs a standard z-score transformation.",
           "The variable's mean value is subtracted from each value, and each is then divided",
           "by the standard deviation. The resulting variable will have a mean of 0 and a",
           "standard deviation of 1. The new variable will have a prefix of RRC_."))
}

on_help_transform_scale01_activate <- function(action, window)
{
  showHelp(Rtxt("Scale [0-1] maps the variable into the 0-1 range.",
           "The new variable will have a prefix of R01_."))
}

on_help_transform_medianmad_activate <- function(action, window)
{
  showHelp(Rtxt("-Median/MAD is a robust version of the standard z-score transform.",
           "The median value is subtracted from each value, and each is then",
           "divided by the median absolute deviation, which is basically the median of",
           "the residuals",
           "or deviations from the data's median, as in Xi-median(X). The resulting",
           "variable will have a median of 0.",
           "It is more resilient to outliers than the normal z-score.",
           "The new variable will have a prefix of RMD_."))
}

on_help_transform_rank_activate <- function(action, window)
{
  showHelp(Rtxt("The Rank transformation will sort the unique numeric values,",
           "and then assign a rank instead of the value. A rank starts from 1 for the minimum",
           "value and increases by 1 for each value up to the maximum value."))
}

on_help_transform_log_activate <- function(action, window)
{
  showHelp(Rtxt("The Log transformation maps the variable using the log function."))
}

on_help_transform_nolan_activate <- function(action, window)
{
  showHelp(Rtxt("The Nolan Groups transformation segments the selected",
           "numeric variables",
           "by a selected categoric variable, and then within each segment rescales the numeric",
           "variable's range to the 0-100 range, using the range option of the rescale(rehsape)",
           "function. This transform was proposed by Anthony Nolan."))
}

on_help_transform_matrix_activate <- function(action, window)
{
  showHelp(Rtxt("The Matrix transformation simply treats the chosen numeric columns",
           "as a matrix. The total sum of the matrix is determined and each value is then",
           "divided by the total."))
}

#------------------------------------------------------------------------

on_help_transform_impute_zero_activate <- function(action, window)
{
  showHelp(Rtxt("Imputation is used to fill in the missing values in the dataset.",
           "The Zero/Missing imputation is a very simple method.",
           "It uses a constant value (0 for numeric data",
           "and the new level, 'missing' for categoric data)",
           "to replace each missing value in the selected variable(s).",
           "<<>>",
           "The imputation of 0 is a good choice if the missing values are likely to",
           "indicate a 0 rather than being unknown. Otherwise, the imputation of 0",
           "may be misleading and is not recommended as it could change the",
           "variables distribution, and hence result in poor models."))
}

on_help_transform_impute_mean_activate <- function(action, window)
{
  showHelp(Rtxt("Imputation is used to fill in the missing values in the data.",
           "The Mean imputation uses the mean (i.e., the average value)",
           "of the variable to replace each missing value in the selected variable(s).",
           "<<>>",
           "Note that this kind of imputation is not recommended as it could change the",
           "variables distribution, and hence result in poor models."))
}

on_help_transform_impute_median_activate <- function(action, window)
{
  showHelp(Rtxt("Imputation is used to fill in the missing values in the dataset.",
           "The Median imputation uses the median (the middle value)",
           "of the variable to replace each missing value in the selected variable(s).",
           "<<>>",
           "Note that this kind of imputation is not recommended as it could change the",
           "variables distribution, and hence result in poor models."))
}

on_help_transform_impute_mode_activate <- function(action, window)
{
  showHelp(Rtxt("Imputation is used to fill in the missing values in the dataset.",
           "The Mode imputation uses the mode (the most frequently occurring value)",
           "of the variable to replace each missing value in the selected variable(s).",
           "<<>>",
           "Note that this kind of imputation is not recommended as it could change the",
           "variable's distribution, and hence result in poor models."))
}


on_help_transform_impute_constant_activate <- function(action, window)
{
  showHelp(Rtxt("Imputation is used to fill in the missing values in the dataset.",
           "The Constant imputation uses the supplied constant value",
           "to replace each missing value in the selected variable(s).",
           "<<>>",
           "Generally, imputing",
           "to a constant value, whether it is zero, the mean, median, or mode, or a user",
           "specified value, is not recommended as it could change the",
           "variables distribution, and hence result in poor models."))
}

#------------------------------------------------------------------------

on_help_transform_remap_quantiles_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Binning based on quantiles will divide the population up",
                   "into nearly equally sized groups.")))
  {
    popupTextviewHelpWindow("quantile")
  }
}

on_help_transform_remap_equal_activate <- function(action, window)
{
  showHelp(Rtxt("Binning based on partitioning the numbers between the minimum and the",
           "maximum value of the variable into equal width segments."))
}

on_help_transform_remap_kmeans_activate <- function(action, window)
{
  showHelp(Rtxt("Binning based on using kmeans clustering will divide the population up",
           "into groups based on distances."))
}

on_help_transform_remap_indicator_activate <- function(action, window)
{
  showHelp(Rtxt("Convert a categoric variable into a collection of indicator variables,",
           "with each new variable corresponding to one of the levels of the categoric variable.",
           "The resulting set of indicator variables each has one of two possible values, 0 or 1.",
           "Only one of the indicator variables, for a particular observation, will have a 1,",
           "and all the rest are necessarily 0. The 1 corresponds to the actual value of the",
           "original variable."))
}

on_help_transform_remap_joincat_activate <- function(action, window)
{
  showHelp(Rtxt("Join two or more categoric variables into a single new categoric variable",
           "which has a new level for each possible value of the contributing levels.",
           "Thus, joining Sex (Male and Female) with AgeGroup (Young, Middle, Old) will result in",
           "JN_Sex_AgeGroup with levels MaleYoung, MaleMiddle, MaleOld, FemaleYoung,",
           "FemaleMiddle, and FemaleOld."))
}

on_help_transform_remap_ascat_activate <- function(action, window)
{
  showHelp(Rtxt("Convert a numeric to a categoric, by turning each distinct value",
           "of the numeric variable in a level for a categoric variable."))
}


on_help_transform_remap_asnum_activate <- function(action, window)
{
  showHelp(Rtxt("Convert a categoric to a numeric, by replacing each level with",
           "the numeric index of the level."))
}

#-----------------------------------------------------------------------

on_help_transform_cleanup_ignored_activate <- function(action, window)
{
  showHelp(Rtxt("Remove any variable that is marked as Ignore."))
}

on_help_transform_cleanup_selected_activate <- function(action, window)
{
  showHelp(Rtxt("Remove all selected variables."))
}

on_help_transform_cleanup_missing_activate <- function(action, window)
{
  showHelp(Rtxt("Remove any variable that has any missing values."))
}

on_help_transform_cleanup_emissing_activate <- function(action, window)
{
  showHelp(Rtxt("Remove any observations (rows) that have any missing values."))
}

########################################################################
# CLUSTER TAB

on_help_kmeans_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("KMeans is a traditional approach to clustering.",
                   "In addition to building a cluster, a discriminant coordinates plot",
                   "can be generated, using the package cluster, as a display of the clusters.")))
  {
    popupTextviewHelpWindow("kmeans")
    if (packageIsAvailable("cluster", viewDocMsg("clusplot")))
    {
      popupTextviewHelpWindow("clusplot", "cluster")
    }
  }
}

on_help_cluster_hclust_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("A hierarchical cluster is built as a hierarchy of clusters,",
                        "starting with each observation defining its own cluster.",
                        "The two closest clusters are then",
                        "combined to generate a new cluster.",
                        "This is repeated until all observations are in the",
                        "one cluster. Thus a hierarchy is generated.",
                        "<<>>",
                        "A complete clustering is built. After building,",
                        "we can choose a specific number of",
                        "clusters that appear 'correct' by viewing the dendrogram.",
                        "Statistics about the chosen",
                        "number of clusters can then be produced.")))
  {
    popupTextviewHelpWindow("hclust")
  }
}

on_help_cluster_stats_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Numerous statistics have been developed over the",
                        "years to measure the performance of a clustering.",
                        "Use the Stats button to report various measures.")))
  {
    popupTextviewHelpWindow("cluster.stats", "fpc")
  }
}

on_help_associate_menuitem_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("Association rule mining finds interesting relationships",
                        "among data items in large data sets.  A typical and widely",
                        "used example of association rule mining is Market Basket",
                        "Analysis, in which customer purchases are analyzed to",
                        "determine what items are frequently purchased together.",
                        "This analysis can guide decisions about the placement of",
                        "items. For example, items that are frequently purchased",
                        "together might be displayed on the same shelf.",
                        "<<>>",
                        "Association rules provide information in the form of",
                        "'if-then' statements. For example, if cookies are",
                        "purchased, milk is going to be purchased, too, with 80",
                        "percent confidence. These rules are probabilistic in",
                        "nature, as the frequencies are computed from the data.")))

    if (packageIsAvailable("arules", viewDocMsg("apriori")))
    {
      popupTextviewHelpWindow("apriori", "arules")
    }
}


########################################################################
# MODEL TAB

on_help_glm_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("A tradition approach to model building is",
                   "regression. Logistic regression (using the binomial family) is used",
                   "to model binary outcomes. Linear regression (using the gaussian family)",
                   "is used to model a linear numeric outcome. For predicting where the",
                   "outcome is a count, the poisson family is used. Further families are",
                   "available, but for now require you to run the glm command directly.",
                   "Please see the additional documentation.",
                   "<<>>",
                   "The R function glm() is used for regression.")))
  {
    popupTextviewHelpWindow("glm")
    popupTextviewHelpWindow("family")
  }
}

on_help_support_vector_machine_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("SVM (Support Vector Machine) is a modern approach",
                   "to modeling where the data is mapped to a higher dimensional space so",
                   "that it is more likely that we can find vectors separating the classes.",
                   "Rattle deploys ksvm from the kernlab package.")))
  {
    #if (packageIsAvailable("e1071", viewDocmsg("e1071"))
    #{
    #  popupTextviewHelpWindow("svm", "e1071")
    #}
    if (packageIsAvailable("kernlab", viewDocMsg("ksvm")))
    {
      popupTextviewHelpWindow("ksvm", "kernlab")
    }
  }
}

on_help_model_nnet_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("A Neural Network is quite an old approach to modeling.",
                   "The structure used for modeling imitates a human's neural network. The idea",
                   "is to build a network of neurons connected by synapses, and instead of",
                   "propagating electrical signals, the network propagates numbers. Mathematically,",
                   "this can be described quite nicely in a relatively straightforward, if not simple,",
                   "way. Rattle uses the functionality provided by the nnet package.")))
  {
    if (packageIsAvailable("nnet", viewDocMsg("nnet")))
    {
      popupTextviewHelpWindow("nnet", "nnet")
    }
  }
}

on_help_model_survival_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("A Survival Model is used to analyze time-to-event data and to generate",
                        "estimated survival curves that show how the probability of the event to",
                        "occur changes over time. In manufacturing, survival analysis is used to",
                        "estimate the time to failure of components and parts; in healthcare it is used",
                        "to estimate the survival probability of patients; in various participation",
                        "programs it is used to estimate the expected time a person will participate in a",
                        "program.   COXPH Survival Analysis will estimate the risk of an event occurring",
                        "for a particular observation relative to the risk for the observed population.",
                        "Parametric Survival Analysis will estimate the expected time within which",
                        "the event will occur.")))
  {
    if (packageIsAvailable("survival", viewDocMsg("coxph")))
    {
      popupTextviewHelpWindow("coxph", "survival")
    }
  }
}

on_help_confusion_table_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("An error (or confusion) matrix concisely reports the performance",
                        "of a model against a testing dataset. Generally, the number of observations",
                        "predicted by the model into each of the classes is presented against the",
                        "actual class to which that observation belongs.",
                        "Rattle reports two error matrices.",
                        "The first is the raw observation counts whilst the second reports the",
                        "percentages.")))
  {
    popupTextviewHelpWindow("table")
  }
}

on_help_risk_chart_activate <- function(action, window)
{
  showHelp(Rtxt("A risk chart plots population proportion along the X axis and a",
           "performance measure along the Y axis. If a Risk variable is identified in the",
           "Data tab then the amount of risk covered is included in the chart (the red line).",
           "The green line indicates the proportion of known targets that are identified",
           "for any proportion of the population covered. The diagonal black line is a baseline,",
           "indicating performance if cases were selected randomly."))
}

on_help_cost_curve_activate <- function(action, window)
{
  showHelp(Rtxt("A cost curve plots the probability cost function against the",
           "normalized expected cost for a range of possible thresholds for the family of",
           "models."))
}

on_help_hand_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("David Hand has developed 4 visualisations to evaluate the",
                        "performance of a model. The hmeasure package implements this and",
                        "is used here to generate the so-called Hand plot.")))
  {
    if (packageIsAvailable("hmeasure", viewDocMsg("HMeasure")))
    {
      popupTextviewHelpWindow("HMeasure", "hmeasure")
    }
  }
}

on_help_lift_activate <- function(action, window)
{
  showHelp(Rtxt("A lift chart plots the lift in performance that is obtained",
           "from different thresholds for the model predicting 0/1."))
}

on_help_roc_activate <- function(action, window)
{
  showHelp(Rtxt("An ROC curve plots the true positives against the false positives."))
}

on_help_precision_activate <- function(action, window)
{
  showHelp(Rtxt("A precision chart plots recall against precision."))
}

on_help_prvob_activate <- function(action, window)
{
  showHelp(Rtxt("The Pr v Ob chart plots the predicted values against the observed values."))
}

on_help_score_activate <- function(action, window)
{
  showHelp(Rtxt("The Score option will apply the selected models to a dataset, saving the",
           "results to file."))
}

on_help_sensitivity_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The Sensitivity versus Specificity chart",
                   "is simply an alternative ROC curve, with Sensitivity being the",
                   "true positive rate (the count of true positives divided by the",
                   "count of positives) and Specificity being the true negative rate",
                   "(the count of true negatives divided by the count of negatives).",
                   "An ROC curve has the false positive rate instead of Specificity, which",
                   "is simply the count of false positives divided by the number of negatives",
                   "(1-fnr).")))
  {
    popupTextviewHelpWindow("performance", "ROCR")
  }
}

on_help_log_activate <- function(action, window)
{
  if (showHelpPlus(Rtxt("The Log tab records the underlying commands that",
                        "Rattle generates and passes to R for execution. This allows",
                        "you to learn about the underlying R commands, and supports",
                        "documentation and repeatability.",
                        "<<>>",
                        "You can export the Log commands to a file and then run that file as",
                        "an R script file at a later stage. The R 'source()' command is used",
                        "to execute an R script file.",
                        "<<>>",
                        "You can also select specific commands and paste them into the attached",
                        "R console, and then modify the command, perhaps to add further",
                        "options that Rattle does not currently support, or simply to experiment",
                        "with directly interacting with R."),
                   extra=Rtxt("Would you like to view the R help on",
                     "the 'source' command?")))
    {
      popupTextviewHelpWindow("source")
    }
}