cute_little_R_functions.R

################################################################
##                                                            ##
##     CUTE FUNCTIONS                                         ##
##                                                            ##
##     Gael A. Millot                                         ##
##                                                            ##
##                                                            ##
################################################################


# color palette: see https://github.com/EmilHvitfeldt/r-color-palettes
# https://usethis.r-lib.org/ and usethat also
## LAST ROUND OF FORMATTING:
# 1) Arguments: one per line
# 2) Description:
# first capital letter everywhere
# AIM
# WARNINGS
# ARGUMENTS
# Update all argument description, saying, character vector, etc, as in fun_gg_boxplot
# RETURN
# Update as in fun_gg_boxplot
# REQUIRED PACKAGES
# REQUIRED FUNCTIONS FROM THE cute PACKAGE
# EXAMPLE
# one example of the example cheet
# see http
# DEBUGGING
# one updated
# 3) Categ (see fun_gg_boxplot for inside code)
# function name
# must have arg.user.setting
# end function name

# required function checking
# see boxplot
# end required function checking

# reserved words (to avoid bugs)
# end reserved words (to avoid bugs)

# arg with no default values
# see boxplot
# end arg with no default values

# argument primary checking
# end argument primary checking

# second round of checking and data preparation
# management of NA arguments
# must have arg.user.setting
# end management of NA arguments

# management of NULL arguments
# end management of NULL arguments

# code that protects set.seed() in the global environment
# end code that protects set.seed() in the global environment

# warning initiation
# end warning initiation

# other checkings
# end other checkings

# reserved word checking
# end reserved word checking
# end second round of checking and data preparation

# package checking
# end package checking
# main code
# output
# if(warn.print == TRUE & ! is.null(warn)){
# end output
# end main code
# 4) example sheet as in fun_gg_boxplot
# 5) test the function with debugging_tools_for_r_dev
# 6) use fun_test()
# 7) write at the beginning of the function:
# todo list check OK
# Check r_debugging_tools-v1.4.R OK
# Check fun_test() (see cute_checks.docx) OK
# example sheet OK 
# check all and any OK # in if() notably, that does not like NA as a logical result
# clear to go Apollo
# 8) check all(, na.rm = TRUE) and any(, na.rm = TRUE)

# fun_mat_fill does not recognize half matrix anymore
# package: 
# Templates: https://prettydoc.statr.me/themes.html
# http://r-pkgs.had.co.nz/
# https://pkgdown.r-lib.org/
# https://rdrr.io/github/gastonstat/cointoss/
# doc:https://www.sphinx-doc.org/en/master/man/sphinx-autogen.html considering that https://www.ericholscher.com/blog/2014/feb/11/sphinx-isnt-just-for-python/
# https://docs.readthedocs.io/en/stable/intro/getting-started-with-sphinx.html
# https://docs.gitlab.com/ee/user/project/pages/
# also register into biotools

# change everywhere: if( ! is.null(arg.check)){

# For heatmap: see https://bioinfo-fr.net/creer-des-heatmaps-a-partir-de-grosses-matrices-en-r

# New function: fun_pdf_conc for Rosine code?
# https://stackoverflow.com/questions/17552917/merging-existing-pdf-files-using-r
# https://www.r-bloggers.com/2019/04/join-split-and-compress-pdf-files-with-pdftools/
# https://rdrr.io/cran/staplr/man/staple_pdf.html

################################ OUTLINE ################################


################ Object analysis    2
######## fun_check() #### check class, type, length, etc., of objects   2
######## fun_secu() #### verif that local variables are not present in other envs   11
######## fun_info() #### recover object information 13
######## fun_head() #### head of the left or right of big 2D objects    15
######## fun_tail() #### tail of the left or right of big 2D objects    16
######## fun_comp_1d() #### comparison of two 1D datasets (vectors, factors, 1D tables) 17
######## fun_comp_2d() #### comparison of two 2D datasets (row & col names, dimensions, etc.)   22
######## fun_comp_list() #### comparison of two lists   29
######## fun_test() #### test combinations of argument values of a function and return errors (and graphs)  32
################ Object modification    47
######## fun_name_change() #### check a vector of character strings and modify any string if present in another vector  47
######## fun_df_remod() #### remodeling a data frame to have column name as a qualitative values and vice-versa 48
######## fun_round() #### rounding number if decimal present    51
######## fun_mat_rotate() #### 90° clockwise matrix rotation    53
######## fun_mat_num2color() #### convert a numeric matrix into hexadecimal color matrix    54
######## fun_mat_op() #### assemble several matrices with operation 58
######## fun_mat_inv() #### return the inverse of a square matrix   60
######## fun_mat_fill() #### fill the empty half part of a symmetric square matrix  62
######## fun_permut() #### progressively breaks a vector order  65
######## fun_slide() #### return a computation made on a vector using a sliding window  76
################ Graphics management    85
######## fun_width() #### window width depending on classes to plot 85
######## fun_open() #### open a GUI or pdf graphic window   87
######## fun_prior_plot() #### set graph param before plotting (erase axes for instance)    91
######## fun_scale() #### select nice label numbers when setting number of ticks on an axis 95
######## fun_inter_ticks() #### define coordinates of secondary ticks   100
######## fun_post_plot() #### set graph param after plotting (axes redesign for instance)   105
######## fun_close() #### close specific graphic windows    117
################ Standard graphics  118
######## fun_empty_graph() #### text to display for empty graphs    118
################ gg graphics    120
######## fun_gg_palette() #### ggplot2 default color palette    120
######## fun_gg_just() #### ggplot2 justification of the axis labeling, depending on angle  122
######## fun_gg_get_legend() #### get the legend of ggplot objects  127
######## fun_gg_point_rast() #### ggplot2 raster scatterplot layer  129
######## fun_gg_boxplot() #### ggplot2 boxplot + background dots if required    133
######## fun_gg_scatter() #### ggplot2 scatterplot + lines (up to 6 overlays totally)   133
######## fun_gg_heatmap() #### ggplot2 heatmap + overlaid mask if required  133
######## fun_gg_empty_graph() #### text to display for empty graphs 141
################ Graphic extraction 143
######## fun_trim() #### display values from a quantitative variable and trim according to defined cut-offs 143
######## fun_segmentation() #### segment a dot cloud on a scatterplot and define the dots from another cloud outside the segmentation   152
################ Import 187
######## fun_pack() #### check if R packages are present and import into the working environment    187
######## fun_python_pack() #### check if python packages are present    189
################ Print / Exporting results (text & tables)  192
######## fun_report() #### print string or data object into output file 192
######## fun_get_message() #### return error/warning/other messages of an expression (that can be exported) 195


################################ FUNCTIONS ################################


################ Object analysis


######## fun_check() #### check class, type, length, etc., of objects


# todo list check OK
# Check r_debugging_tools-v1.4.R OK
# Check fun_test() 20201107 (see cute_checks.docx) OK
# example sheet OK 
# check all and any OK
# -> clear to go Apollo
fun_check <- function(
data, 
class = NULL, 
typeof = NULL, 
mode = NULL, 
length = NULL, 
prop = FALSE, 
double.as.integer.allowed = FALSE, 
options = NULL, 
all.options.in.data = FALSE, 
na.contain = FALSE, 
neg.values = TRUE, 
print = FALSE, 
data.name = NULL, 
fun.name = NULL
){
# AIM
# Check the class, type, mode and length of the data argument
# Mainly used to check the arguments of other functions
# Check also other kind of data parameters, is it a proportion? Is it type double but numbers without decimal part?
# If options == NULL, then at least class or type or mode or length argument must be non-null
# If options is non-null, then class, type and mode must be NULL, and length can be NULL or specified
# WARNINGS
# The function tests what is written in its arguments, even if what is written is incoherent. For instance, fun_check(data = factor(1), class = "factor", mode = "character") will return a problem, whatever the object tested in the data argument, because no object can be class "factor" and mode "character" (factors are class "factor" and mode "numeric"). Of note, length of object of class "environment" is always 0
# If the tested object is NULL, then the function will always return a checking problem
# Since R >= 4.0.0, class(matrix()) returns "matrix" "array", and not "matrix" alone as before. However, use argument class = "matrix" to check for matrix object (of class "matrix" "array" in R >= 4.0.0) and use argument class = "array" to check for array object (of class "array" in R >= 4.0.0)
# ARGUMENTS
# data: object to test
# class: character string. Either one of the class() result (But see the warning section above) or "vector" or "ggplot2" (i.e., objects of class c("gg", "ggplot")) or NULL
# typeof: character string. Either one of the typeof() result or NULL
# mode: character string. Either one of the mode() result (for non-vector object) or NULL
# length: numeric value indicating the length of the object. Not considered if NULL
# prop: logical. Are the numeric values between 0 and 1 (proportion)? If TRUE, can be used alone, without considering class, etc.
# double.as.integer.allowed: logical. If TRUE, no error is reported in the cheking message if argument is set to typeof == "integer" or class == "integer", while the reality is typeof == "double" or class == "numeric" but the numbers strictly have zero as modulo (remainder of a division). This means that i <- 1, which is typeof(i) -> "double" is considered as integer with double.as.integer.allowed = TRUE. WARNING: data%%1 == 0 but not isTRUE(all.equal(data%%1, 0)) is used here because the argument checks for integers stored as double (does not check for decimal numbers that are approximate integers)
# options: a vector of character strings indicating all the possible option values for the data argument, or NULL
# all.options.in.data: logical. If TRUE, all of the options must be present at least once in the data argument, and nothing else. If FALSE, some or all of the options must be present in the data argument, and nothing else. Ignored if options is NULL
# na.contain: logical. Can the data argument contain NA?
# neg.values: logical. Are negative numeric values authorized? Warning: the default setting is TRUE, meaning that, in that case, no check is performed for the presence of negative values. The neg.values argument is activated only when set to FALSE. In addition, (1) neg.values = FALSE can only be used when class, typeof or mode arguments are not NULL, otherwise return an error message, (2) the presence of negative values is not checked with neg.values = FALSE if the tested object is a factor and the following checking message is returned "OBJECT MUST BE MADE OF NON NEGATIVE VALUES BUT IS A FACTOR"
# print: logical. Print the message if $problem is TRUE? Warning: set by default to FALSE, which facilitates the control of the checking message output when using fun_check() inside functions. See the example section
# data.name: character string indicating the name of the object to test. If NULL, use what is assigned to the data argument for the returned message
# fun.name: character string indicating the name of the function checked (i.e., when fun_check() is used to check the arguments of this function). If non-null, the value of fun.name will be added into the message returned by fun_check()
# RETURN
# A list containing:
# $problem: logical. Is there any problem detected?
# $text: message indicating the details of the problem, or the absence of problem
# $object.name: value of the data.name argument (i.e., name of the checked object if provided, NULL otherwise)
# REQUIRED PACKAGES
# None
# REQUIRED FUNCTIONS FROM THE cute PACKAGE
# None
# EXAMPLE
# test <- matrix(1:3) ; fun_check(data = test, print = TRUE, class = "vector", mode = "numeric")
# see http
# DEBUGGING
# data = mean ; class = NULL ; typeof = NULL ; mode = NULL ; length = NULL ; prop = FALSE ; double.as.integer.allowed = FALSE ; options = "a" ; all.options.in.data = FALSE ; na.contain = FALSE ; neg.values = TRUE ; print = TRUE ; data.name = NULL ; fun.name = NULL
# function name
# no used in this function for the error message, to avoid env colliding
# end function name
# required function checking
# end required function checking
# reserved words
# end reserved words
# fun.name checked first because required next
if( ! is.null(fun.name)){ # I have to use this way to deal with every kind of class for fun.name
if(all(base::class(fun.name) == "character")){ # all() without na.rm -> ok because class(NA) is "logical"
if(base::length(fun.name) != 1){
tempo.cat <- paste0("ERROR IN fun_check(): THE fun.name ARGUMENT MUST BE A CHARACTER VECTOR OF LENGTH 1: ", paste(fun.name, collapse = " "))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}else if(any(is.na(fun.name))){ # normally no NA with is.na()
tempo.cat <- paste0("ERROR IN fun_check(): NO ARGUMENT EXCEPT data AND options CAN HAVE NA VALUES\nPROBLEMATIC ARGUMENT IS fun.name")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}else{
tempo.cat <- paste0("ERROR IN fun_check(): THE fun.name ARGUMENT MUST BE A CHARACTER VECTOR OF LENGTH 1") # paste(fun.name, collapse = " ") removed here because does not work with objects like function
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
# end fun.name checked first because required next
# arg with no default values
if(missing(data)){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" IN ", fun.name)), ": ARGUMENT data HAS NO DEFAULT VALUE AND REQUIRES ONE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end arg with no default values
# argument primary checking
# source("C:/Users/Gael/Documents/Git_versions_to_use/debugging_tools_for_r_dev-v1.7/r_debugging_tools-v1.7.R") ; eval(parse(text = str_basic_arg_check_dev)) # activate this line and use the function to check arguments status
# end argument primary checking
# second round of checking and data preparation
# management of special classes
basic.class <- c(
"NULL", # because class(NULL) is "NULL". The NULL aspect will be dealt later
"logical", 
"integer", 
"numeric", 
# "complex", 
"character"
# "matrix", 
# "array", 
# "data.frame", 
# "list", 
# "factor", 
# "table", 
# "expression", 
# "name", 
# "symbol", 
# "function", 
# "uneval", 
# "environment", 
# "ggplot2", 
# "ggplot_built", 
# "call"
)
tempo.arg.base <-c( # no names(formals(fun = sys.function(sys.parent(n = 2)))) used with fun_check() to be sure to deal with the correct environment
"class", 
"typeof", 
"mode", 
"length", 
"prop", 
"double.as.integer.allowed", 
"options", 
"all.options.in.data", 
"na.contain", 
"neg.values", 
"print", 
"data.name", 
"fun.name"
)
tempo.class <-list( # no get() used to be sure to deal with the correct environment
base::class(class), 
base::class(typeof), 
base::class(mode), 
base::class(length), 
base::class(prop), 
base::class(double.as.integer.allowed), 
base::class(options), 
base::class(all.options.in.data), 
base::class(na.contain), 
base::class(neg.values), 
base::class(print), 
base::class(data.name), 
base::class(fun.name)
)
tempo <- ! sapply(lapply(tempo.class, FUN = "%in%", basic.class), FUN = all)
if(any(tempo)){
tempo.cat1 <- tempo.arg.base[tempo]
tempo.cat2 <- sapply(tempo.class[tempo], FUN = paste0, collapse = " ")
tempo.sep <- sapply(mapply(" ", max(nchar(tempo.cat1)) - nchar(tempo.cat1) + 3, FUN = rep, SIMPLIFY = FALSE), FUN = paste0, collapse = "")
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": ANY ARGUMENT EXCEPT data MUST HAVE A BASIC CLASS\nPROBLEMATIC ARGUMENT", ifelse(base::length(tempo.cat1) > 1, "S", ""), " AND ASSOCIATED CLASS", ifelse(base::length(tempo.cat1) > 1, "ES ARE", " IS"), ":\n", paste0(tempo.cat1, tempo.sep, tempo.cat2, collapse = "\n")) # normally no NA with is.na()
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end management of special classes
# management of NA arguments
if(any(is.na(data.name)) | any(is.na(class)) | any(is.na(typeof)) | any(is.na(mode)) | any(is.na(length)) | any(is.na(prop)) | any(is.na(double.as.integer.allowed)) | any(is.na(all.options.in.data)) | any(is.na(na.contain)) | any(is.na(neg.values)) | any(is.na(print)) | any(is.na(fun.name))){ # normally no NA with is.na()
tempo <- c("data.name", "class", "typeof", "mode", "length", "prop", "double.as.integer.allowed", "all.options.in.data", "na.contain", "neg.values", "print", "fun.name")[c(any(is.na(data.name)), any(is.na(class)), any(is.na(typeof)), any(is.na(mode)), any(is.na(length)), any(is.na(prop)), any(is.na(double.as.integer.allowed)), any(is.na(all.options.in.data)), any(is.na(na.contain)), any(is.na(neg.values)), any(is.na(print)), any(is.na(fun.name)))]
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": NO ARGUMENT EXCEPT data AND options CAN HAVE NA VALUES\nPROBLEMATIC ARGUMENT", ifelse(length(tempo) > 1, "S ARE", " IS"), ":\n", paste(tempo, collapse = "\n")) # normally no NA with is.na()
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end management of NA arguments
# management of NULL arguments
tempo.arg <-c(
"prop", 
"double.as.integer.allowed", 
"all.options.in.data", 
"na.contain",
"neg.values",
"print"
)
tempo.log <- sapply(lapply(tempo.arg, FUN = get, env = sys.nframe(), inherit = FALSE), FUN = is.null)
if(any(tempo.log) == TRUE){ # normally no NA with is.null()
tempo.cat <- paste0("ERROR IN fun.check():\n", ifelse(sum(tempo.log, na.rm = TRUE) > 1, "THESE ARGUMENTS", "THIS ARGUMENT"), " CANNOT BE NULL:\n", paste0(tempo.arg[tempo.log], collapse = "\n"))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end management of NULL arguments
# dealing with logical
# tested below
# end dealing with logical
# code that protects set.seed() in the global environment
# end code that protects set.seed() in the global environment
# warning initiation
# end warning initiation
# other checkings
if( ! is.null(data.name)){
if( ! (base::length(data.name) == 1 & all(base::class(data.name) == "character"))){ # all() without na.rm -> ok because class(NA) is "logical"
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": data.name ARGUMENT MUST BE A SINGLE CHARACTER ELEMENT AND NOT ", paste(data.name, collapse = " "))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
if(is.null(options) & is.null(class) & is.null(typeof) & is.null(mode) &  prop == FALSE & is.null(length)){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": AT LEAST ONE OF THE options, class, typeof, mode, prop, OR length ARGUMENT MUST BE SPECIFIED (I.E, TRUE FOR prop)")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if( ! is.null(options) & ( ! is.null(class) | ! is.null(typeof) | ! is.null(mode) | prop == TRUE)){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": THE class, typeof, mode ARGUMENTS MUST BE NULL, AND prop FALSE, IF THE options ARGUMENT IS SPECIFIED\nTHE options ARGUMENT MUST BE NULL IF THE class AND/OR typeof AND/OR mode AND/OR prop ARGUMENT IS SPECIFIED")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if( ! (all(base::class(neg.values) == "logical") & base::length(neg.values) == 1)){ # all() without na.rm -> ok because class(NA) is "logical" 
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": THE neg.values ARGUMENT MUST BE TRUE OR FALSE ONLY")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if(neg.values == FALSE & is.null(class) & is.null(typeof) & is.null(mode)){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": THE neg.values ARGUMENT CANNOT BE SWITCHED TO FALSE IF class, typeof AND mode ARGUMENTS ARE NULL")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if( ! is.null(class)){ # may add "formula" and "Date" as in https://renenyffenegger.ch/notes/development/languages/R/functions/class
if( ! all(class %in% c("vector", "logical", "integer", "numeric", "complex", "character", "matrix", "array", "data.frame", "list", "factor", "table", "expression", "name", "symbol", "function", "uneval", "environment", "ggplot2", "ggplot_built", "call") & base::length(class) == 1)){ # length == 1 here because of class(matrix()) since R4.0.0  # all() without na.rm -> ok because class cannot be NA (tested above)
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": class ARGUMENT MUST BE ONE OF THESE VALUE:\n\"vector\", \"logical\", \"integer\", \"numeric\", \"complex\", \"character\", \"matrix\", \"array\", \"data.frame\", \"list\", \"factor\", \"table\", \"expression\", \"name\", \"symbol\", \"function\", \"environment\", \"ggplot2\", \"ggplot_built\", \"call\"")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if(neg.values == FALSE & ! any(class %in% c("vector", "numeric", "integer", "matrix", "array", "data.frame", "table"))){ # no need of na.rm = TRUE for any() because %in% does not output NA
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": class ARGUMENT CANNOT BE OTHER THAN \"vector\", \"numeric\", \"integer\", \"matrix\", \"array\", \"data.frame\", \"table\" IF neg.values ARGUMENT IS SWITCHED TO FALSE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
if( ! is.null(typeof)){ # all the types are here: https://renenyffenegger.ch/notes/development/languages/R/functions/typeof
if( ! (all(typeof %in% c("logical", "integer", "double", "complex", "character", "list", "expression", "symbol", "closure", "special", "builtin", "environment", "S4", "language")) & base::length(typeof) == 1)){ # "language" is the type of object of class "call" # all() without na.rm -> ok because typeof cannot be NA (tested above)
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": typeof ARGUMENT MUST BE ONE OF THESE VALUE:\n\"logical\", \"integer\", \"double\", \"complex\", \"character\", \"list\", \"expression\", \"name\", \"symbol\", \"closure\", \"special\", \"builtin\", \"environment\", \"S4\", \"language\"")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if(neg.values == FALSE & ! typeof %in% c("double", "integer")){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": typeof ARGUMENT CANNOT BE OTHER THAN \"double\" OR \"integer\" IF neg.values ARGUMENT IS SWITCHED TO FALSE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
if( ! is.null(mode)){ # all the types are here: https://renenyffenegger.ch/notes/development/languages/R/functions/typeof
if( ! (all(mode %in% c("logical", "numeric", "complex", "character", "list", "expression", "name", "symbol", "function", "environment", "S4", "call")) & base::length(mode) == 1)){ # all() without na.rm -> ok because mode cannot be NA (tested above)
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": mode ARGUMENT MUST BE ONE OF THESE VALUE:\n\"logical\", \"numeric\", \"complex\", \"character\", \"list\", \"expression\", \"name\", \"symbol\", \"function\", \"environment\", \"S4\", \"call\"")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if(neg.values == FALSE & mode != "numeric"){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": mode ARGUMENT CANNOT BE OTHER THAN \"numeric\" IF neg.values ARGUMENT IS SWITCHED TO FALSE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
if( ! is.null(length)){
if( ! (is.numeric(length) & base::length(length) == 1 & ! grepl(length, pattern = "\\."))){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": length ARGUMENT MUST BE A SINGLE INTEGER VALUE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
if( ! (is.logical(prop) & base::length(prop) == 1)){ # is.na() already checked for prop
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": prop ARGUMENT MUST BE TRUE OR FALSE ONLY")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}else if(prop == TRUE){
if( ! is.null(class)){
if( ! any(class %in% c("vector", "numeric", "matrix", "array", "data.frame", "table"))){ # no need of na.rm = TRUE for any() because %in% does not output NA
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": class ARGUMENT CANNOT BE OTHER THAN NULL, \"vector\", \"numeric\", \"matrix\", \"array\", \"data.frame\", \"table\" IF prop ARGUMENT IS TRUE") # not integer because prop
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
if( ! is.null(mode)){
if(mode != "numeric"){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": mode ARGUMENT CANNOT BE OTHER THAN NULL OR \"numeric\" IF prop ARGUMENT IS TRUE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
if( ! is.null(typeof)){
if(typeof != "double"){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": typeof ARGUMENT CANNOT BE OTHER THAN NULL OR \"double\" IF prop ARGUMENT IS TRUE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
}
}
if( ! (all(base::class(double.as.integer.allowed) == "logical") & base::length(double.as.integer.allowed) == 1)){ # all() without na.rm -> ok because class() never returns NA
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": THE double.as.integer.allowed ARGUMENT MUST BE TRUE OR FALSE ONLY: ", paste(double.as.integer.allowed, collapse = " "))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if( ! (is.logical(all.options.in.data) & base::length(all.options.in.data) == 1)){
tempo.cat <- paste0("ERROR IN fun_check()", ifelse(is.null(fun.name), "", paste0(" INSIDE ", fun.name)), ": all.options.in.data ARGUMENT MUST BE A SINGLE LOGICAL VALUE (TRUE OR FALSE ONLY): ", paste(all.options.in.data, collapse = " "))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if( ! (all(base::class(na.contain) == "logical") & base::length(na.contain) == 1)){ # all() without na.rm -> ok because class() never returns NA
tempo.cat <- paste0("ERROR IN fun_check(): THE na.contain ARGUMENT MUST BE TRUE OR FALSE ONLY: ", paste(na.contain, collapse = " "))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
if( ! (all(base::class(print) == "logical") & base::length(print) == 1)){ # all() without na.rm -> ok because class() never returns NA
tempo.cat <- paste0("ERROR IN fun_check(): THE print ARGUMENT MUST BE TRUE OR FALSE ONLY: ", paste(print, collapse = " "))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# data.name and fun.name tested at the beginning
# end other checkings
# end second round of checking and data preparation
# package checking
# end package checking
# main code
if(is.null(data.name)){
data.name <- deparse(substitute(data))
}
problem <- FALSE
text <- paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT")
if(( ! is.null(options)) & all(base::typeof(data) == "character")){ # all() without na.rm -> ok because typeof() never returns NA
text <- ""
if( ! all(data %in% options)){ # no need of na.rm = TRUE for all() because %in% does not output NA
problem <- TRUE
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": THE ", data.name, " OBJECT MUST BE SOME OF THESE OPTIONS: ", paste(options, collapse = " "), "\nTHE PROBLEMATIC ELEMENTS OF ", data.name, " ARE: ", paste(unique(data[ ! (data %in% options)]), collapse = " "))
}
if(all.options.in.data == TRUE){
if( ! all(options %in% data)){ # no need of na.rm = TRUE for all() because %in% does not output NA
problem <- TRUE
text <- paste0(ifelse(text == "", "", paste0(text, "\n")), ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": THE ", data.name, " OBJECT MUST BE MADE OF ALL THESE OPTIONS: ", paste(options, collapse = " "), "\nTHE MISSING ELEMENTS OF THE options ARGUMENT ARE: ",  paste(unique(options[ ! (options %in% data)]), collapse = " "))
}
}
if( ! is.null(length)){
if(base::length(data) != length){
problem <- TRUE
text <- paste0(ifelse(text == "", "", paste0(text, "\n")), ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": THE LENGTH OF ", data.name, " MUST BE ", length, " AND NOT ", base::length(data))
}
}
if(text == ""){
text <- paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT")
}
}else if( ! is.null(options)){
problem <- TRUE
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": THE ", data.name, " OBJECT MUST BE SOME OF THESE OPTIONS: ", paste(options, collapse = " "), "\nBUT IS NOT EVEN TYPE CHARACTER")
}
arg.names <- c("class", "typeof", "mode", "length")
if( ! is.null(class)){
if(class == "matrix"){ # because of class(matric()) since R4.0.0
class <- c("matrix", "array")
}else if(class == "factor" & all(base::class(data) %in% c("factor", "ordered"))){ # to deal with ordered factors # all() without na.rm -> ok because class(NA) is "logical"
class <- c("factor", "ordered")
}
}
if(is.null(options)){
for(i2 in 1:base::length(arg.names)){
if( ! is.null(get(arg.names[i2], env = sys.nframe(), inherit = FALSE))){
# script to execute
tempo.script <- '
problem <- TRUE ;
if(identical(text, paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT"))){
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": THE ", data.name, " OBJECT MUST BE ") ;
}else{
text <- paste0(text, " AND ") ; 
}
text <- paste0(text, toupper(arg.names[i2]), " ", if(all(get(arg.names[i2], env = sys.nframe(), inherit = FALSE) %in% c("matrix", "array"))){"matrix"}else if(all(get(arg.names[i2], env = sys.nframe(), inherit = FALSE) %in% c("factor", "ordered"))){"factor"}else{get(arg.names[i2], env = sys.nframe(), inherit = FALSE)})
' # no need of na.rm = TRUE for all() because %in% does not output NA
# end script to execute
if(base::typeof(data) == "double" & double.as.integer.allowed == TRUE & ((arg.names[i2] == "class" & all(get(arg.names[i2], env = sys.nframe(), inherit = FALSE) == "integer")) | (arg.names[i2] == "typeof" & all(get(arg.names[i2], env = sys.nframe(), inherit = FALSE) == "integer")))){ # no need of na.rm = TRUE for all() because == does not output NA if no NA in left of ==, which is the case for arg.names # typeof(data) == "double" means no factor allowed
if( ! all(data %% 1 == 0, na.rm = TRUE)){ # to check integers (use %%, meaning the remaining of a division): see the precedent line. isTRUE(all.equal(data%%1, rep(0, length(data)))) not used because we strictly need zero as a result. Warning: na.rm = TRUE required here for all()
eval(parse(text = tempo.script)) # execute tempo.script
}
}else if( ! any(all(get(arg.names[i2], env = sys.nframe(), inherit = FALSE) %in% c("vector", "ggplot2"))) & ! all(eval(parse(text = paste0(arg.names[i2], "(data)"))) %in% get(arg.names[i2], env = sys.nframe(), inherit = FALSE))){ # test the four c("class", "typeof", "mode", "length") arguments with their corresponding function. No need of na.rm = TRUE for all() because %in% does not output NA # no need of na.rm = TRUE for all() because %in% does not output NA # no need of na.rm = TRUE for any() because get get(arg.names) does not contain NA
eval(parse(text = tempo.script)) # execute tempo.script
}else if(arg.names[i2] == "class" & all(get(arg.names[i2], env = sys.nframe(), inherit = FALSE) == "vector") & ! (all(base::class(data) %in% "numeric") | all(base::class(data) %in% "integer") | all(base::class(data) %in% "character") | all(base::class(data) %in% "logical"))){ # test class == "vector". No need of na.rm = TRUE for all() because %in% does not output NA # no need of na.rm = TRUE for all() because == does not output NA if no NA in left of ==, which is the case for arg.names
eval(parse(text = tempo.script)) # execute tempo.script
}else if(arg.names[i2] == "class" & all(get(arg.names[i2], env = sys.nframe(), inherit = FALSE) == "ggplot2") & ! all(base::class(data) %in% c("gg", "ggplot"))){ # test ggplot object # no need of na.rm = TRUE for all() because == does not output NA if no NA in left of ==, which is the case for arg.names # no need of na.rm = TRUE for all() because %in% does not output NA
eval(parse(text = tempo.script)) # execute tempo.script
}
}
}
}
if(prop == TRUE & all(base::typeof(data) == "double")){ # all() without na.rm -> ok because typeof(NA) is "logical"
if(is.null(data) | any(data < 0 | data > 1, na.rm = TRUE)){ # works if data is NULL # Warning: na.rm = TRUE required here for any() # typeof(data) == "double" means no factor allowed
problem <- TRUE
if(identical(text, paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT"))){
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": ")
}else{
text <- paste0(text, " AND ")
}
text <- paste0(text, "THE ", data.name, " OBJECT MUST BE DECIMAL VALUES BETWEEN 0 AND 1")
}
}else if(prop == TRUE){
problem <- TRUE
if(identical(text, paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT"))){
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": ")
}else{
text <- paste0(text, " AND ")
}
text <- paste0(text, "THE ", data.name, " OBJECT MUST BE DECIMAL VALUES BETWEEN 0 AND 1")
}
if(all(base::class(data) %in% "expression")){ # no need of na.rm = TRUE for all() because %in% does not output NA
data <- as.character(data) # to evaluate the presence of NA
}
if(na.contain == FALSE & (base::mode(data) %in% c("logical", "numeric", "complex", "character", "list"))){ # before it was ! (class(data) %in% c("function", "environment"))
if(any(is.na(data)) == TRUE){ # not on the same line because when data is class envir or function , do not like that # normally no NA with is.na()
problem <- TRUE
if(identical(text, paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT"))){
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": ")
}else{
text <- paste0(text, " AND ")
}
text <- paste0(text, "THE ", data.name, " OBJECT CONTAINS NA WHILE NOT AUTHORIZED")
}
}
if(neg.values == FALSE & all(base::mode(data) %in% "numeric") & ! any(base::class(data) %in% "factor")){ # no need of na.rm = TRUE for all() because %in% does not output NA
if(any(data < 0, na.rm = TRUE)){ # Warning: na.rm = TRUE required here for any()
problem <- TRUE
if(identical(text, paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT"))){
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": ")
}else{
text <- paste0(text, " AND ")
}
text <- paste0(text, "THE ", data.name, " OBJECT MUST BE MADE OF NON NEGATIVE NUMERIC VALUES")
}
}else if(neg.values == FALSE){
problem <- TRUE
if(identical(text, paste0(ifelse(is.null(fun.name), "", paste0("IN ", fun.name, ": ")), "NO PROBLEM DETECTED FOR THE ", data.name, " OBJECT"))){
text <- paste0(ifelse(is.null(fun.name), "ERROR", paste0("ERROR IN ", fun.name)), ": ")
}else{
text <- paste0(text, " AND ")
}
text <- paste0(text, "THE ", data.name, " OBJECT MUST BE MADE OF NON NEGATIVE VALUES BUT IS ", ifelse(any(base::class(data) %in% "factor"), "A FACTOR", "NOT EVEN MODE NUMERIC"))
}
if(print == TRUE & problem == TRUE){
cat(paste0("\n\n================\n\n", text, "\n\n================\n\n"))
}
# output
output <- list(problem = problem, text = text, object.name = data.name)
return(output)
# end output
# end main code
}

######## fun_secu() #### verif that local variables are not present in other envs


fun_secu <- function(pos = 1, name = NULL){
# AIM
# Verify that local variables are not present in other environments, in order to avoid scope preference usage. The fun_secu() function checks by default the parent environment. This means that when used inside a function, it checks the local environment of this function. When used in the Global environment, it would check this environment
# ARGUMENTS
# pos: single integer indicating the position of the environment checked (argument n of parent.frame()). Value 1 means one step above the fun_secu() local environment. Thus, if fun_secu() is used in the working environment, with pos ==1, variables of the working env will be checked if they also exist in the above envs (Scope of R). If fun_secu() is used in a function, with pos ==1, variables presents in the local env of the function will be checked if they also exist in the above envs (Scope of R, which includes the working environment (Global env))
# name: single character string indicating the name of the function checked. If NULL, fun_secu() checks all the variables of the environment indicated by pos, as explained in the pos argument description. If non-null, fun_secu() checks all the variables presents in the local env of the function will be checked in the above envs (which includes the working environment (Global env)
# RETURN
# A character string of the local variables that match variables in the different environments of the R scope, or NULL if no match
# REQUIRED PACKAGES
# None
# REQUIRED FUNCTIONS FROM CUTE_LITTLE_R_FUNCTION
# fun_check()
# EXAMPLES
# fun_secu()
# fun_secu(pos = 2)
# mean <- 0 ; fun1 <- function(){sd <- 1 ; fun_secu(name = as.character(sys.calls()[[length(sys.calls())]]))} ; fun2 <- function(){cor <- 2 ; fun1()} ; fun1() ; fun2() ; rm(mean) # sys.calls() gives the the function name at top stack of the imbricated functions, sys.calls()[[length(sys.calls())]] the name of the just above function. This can also been used for the above function: as.character(sys.call(1))
# test.pos <- 2 ; mean <- 0 ; fun1 <- function(){sd <- 1 ; fun_secu(pos = test.pos, name = if(length(sys.calls()) >= test.pos){as.character(sys.calls()[[length(sys.calls()) + 1 - test.pos]])}else{search()[ (1:length(search()))[test.pos - length(sys.calls())]]})} ; fun2 <- function(){cor <- 2 ; fun1()} ; fun1() ; fun2() ; rm(mean) # for argument name, here is a way to have the name of the tested environment according to test.pos value
# DEBUGGING
# pos = 1 ; name = NULL # for function debugging
# function name
function.name <- paste0(as.list(match.call(expand.dots = FALSE))[[1]], "()")
arg.user.setting <- as.list(match.call(expand.dots = FALSE))[-1] # list of the argument settings (excluding default values not provided by the user)
# end function name
# required function checking
if(length(utils::find("fun_check", mode = "function")) == 0){
tempo.cat <- paste0("ERROR IN ", function.name, "\nREQUIRED fun_check() FUNCTION IS MISSING IN THE R ENVIRONMENT")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end required function checking
# argument primary checking
# arg with no default values
# end arg with no default values
# using fun_check()
arg.check <- NULL #
text.check <- NULL #
checked.arg.names <- NULL # for function debbuging: used by r_debugging_tools
ee <- expression(arg.check <- c(arg.check, tempo$problem) , text.check <- c(text.check, tempo$text) , checked.arg.names <- c(checked.arg.names, tempo$object.name))
tempo <- fun_check(data = pos, class = "vector", typeof = "integer", double.as.integer.allowed = TRUE, length = 1, fun.name = function.name) ; eval(ee)
if( ! is.null(name)){
tempo <- fun_check(data = name, class = "vector", typeof = "character", length = 1, fun.name = function.name) ; eval(ee)
}
if(any(arg.check) == TRUE){
stop(paste0("\n\n================\n\n", paste(text.check[arg.check], collapse = "\n"), "\n\n================\n\n"), call. = FALSE) #
}
# end using fun_check()
# source("C:/Users/Gael/Documents/Git_versions_to_use/debugging_tools_for_r_dev-v1.7/r_debugging_tools-v1.7.R") ; eval(parse(text = str_basic_arg_check_dev)) ; eval(parse(text = str_arg_check_with_fun_check_dev)) # activate this line and use the function (with no arguments left as NULL) to check arguments status and if they have been checked using fun_check()
# end argument primary checking
# second round of checking and data preparation
# management of NA arguments
tempo.arg <- names(arg.user.setting) # values provided by the user
tempo.log <- suppressWarnings(sapply(lapply(lapply(tempo.arg, FUN = get, env = sys.nframe(), inherit = FALSE), FUN = is.na), FUN = any)) & lapply(lapply(tempo.arg, FUN = get, env = sys.nframe(), inherit = FALSE), FUN = length) == 1 # no argument provided by the user can be just NA
if(any(tempo.log) == TRUE){
tempo.cat <- paste0("ERROR IN ", function.name, "\n", ifelse(sum(tempo.log, na.rm = TRUE) > 1, "THESE ARGUMENTS", "THIS ARGUMENT"), " CANNOT JUST BE NA:", paste0(tempo.arg[tempo.log], collapse = "\n"))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end management of NA arguments
# management of NULL arguments
tempo.arg <- c(
"pos"
)
tempo.log <- sapply(lapply(tempo.arg, FUN = get, env = sys.nframe(), inherit = FALSE), FUN = is.null)
if(any(tempo.log) == TRUE){
tempo.cat <- paste0("ERROR IN ", function.name, "\n", ifelse(sum(tempo.log, na.rm = TRUE) > 1, "THESE ARGUMENTS", "THIS ARGUMENT"), " CANNOT BE NULL:", paste0(tempo.arg[tempo.log], collapse = "\n"))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end management of NULL arguments
# end second round of checking and data preparation
# main code
# match.list <- vector("list", length = (length(sys.calls()) - 1 + length(search()) + ifelse(length(sys.calls()) == 1, -1, 0))) # match.list is a list of all the environment tested (local of functions and R envs), length(sys.calls()) - 1 to remove the level of the fun_secu() function, sys.calls() giving all the names of the imbricated functions, including fun_secu, ifelse(length(sys.calls()) == 1, -1, 0) to remove Global env if this one is tested
tempo.name <- rev(as.character(unlist(sys.calls()))) # get names of frames (i.e., enclosed env)
tempo.frame <- rev(sys.frames())  # get frames (i.e., enclosed env)
# dealing with source()
# source() used in the Global env creates three frames above the Global env, which should be removed because not very interesting for variable duplications. Add a <<-(sys.frames()) in this code and source anova_contrasts code to see this. With ls(a[[4]]), we can see the content of each env, which are probably elements of source()
if(any(sapply(tempo.frame, FUN = environmentName) %in% "R_GlobalEnv")){
global.pos <- which(sapply(tempo.frame, FUN = environmentName) %in% "R_GlobalEnv")
# remove the global env (because already in search(), and all the oabove env
tempo.name <- tempo.name[-c(global.pos:length(tempo.frame))]
tempo.frame <- tempo.frame[-c(global.pos:length(tempo.frame))]
}
# end dealing with source()
# might have a problem if(length(tempo.name) == 0){
match.list <- vector("list", length = length(tempo.name) + length(search())) # match.list is a list of all the environment tested (local of functions and R envs), length(sys.calls()) - 1 to remove the level of the fun_secu() function, sys.calls() giving all the names of the imbricated functions, including fun_secu, ifelse(length(sys.calls()) == 1, -1, 0) to remove Global env if this one is tested
ls.names <- c(tempo.name, search()) # names of the functions + names of the search() environments
ls.input <- c(tempo.frame, as.list(search())) # environements of the functions + names of the search() environments
names(match.list) <- ls.names # 
match.list <- match.list[-c(1:(pos + 1))] # because we check only above pos
ls.tested <- ls.input[[pos + 1]]
ls.input <- ls.input[-c(1:(pos + 1))]
for(i1 in 1:length(match.list)){
if(any(ls(name = ls.input[[i1]], all.names = TRUE) %in% ls(name = ls.tested, all.names = TRUE))){
match.list[i1] <- list(ls(name = ls.input[[i1]], all.names = TRUE)[ls(name = ls.input[[i1]], all.names = TRUE) %in% ls(name = ls.tested, all.names = TRUE)])
}
}
if( ! all(sapply(match.list, FUN = is.null))){
output <- paste0("SOME VARIABLES ", ifelse(is.null(name), "OF THE CHECKED ENVIRONMENT", paste0("OF ", name)), " ARE ALSO PRESENT IN :\n", paste0(names(match.list[ ! sapply(match.list, FUN = is.null)]), ": ", sapply(match.list[ ! sapply(match.list, FUN = is.null)], FUN = paste0, collapse = " "), collapse = "\n"))
}else{
output <- NULL
}
return(output)
}


######## fun_info() #### broad description of an object


# todo list check OK
# Check r_debugging_tools-v1.4.R OK
# Check fun_test() (see cute_checks.docx) OK
# example sheet OK 
# check all and any OK
# -> clear to go Apollo
fun_info <- function(
data, 
n = NULL, 
warn.print = TRUE
){
# AIM
# Provide a broad description of an object
# WARNINGS
# None
# ARGUMENTS
# data: object to analyse
# n: positive integer value indicating the n first number of elements to display per compartment of the output list (i.e., head(..., n)). Write NULL to return all the elements. Does not apply for the $STRUCTURE compartment output
# warn.print: logical. Print potential warnings at the end of the execution? If FALSE the warning messages are added in the output list as an additional compartment (or NULL if no message).
# RETURN
# A list containing information, depending on the class and type of data. The backbone is generally:
# $NAME: name of the object
# $CLASS: class of the object (class() value)
# $TYPE: type of the object (typeof() value)
# $LENGTH: length of the object (length() value)
# $NA.NB: number of NA and NaN (only for type "logical", "integer", "double", "complex", "character" or "list")
# $HEAD: head of the object (head() value)
# $TAIL: tail of the object (tail() value)
# $DIMENSION: dimension (only for object with dimensions)
# $SUMMARY: object summary (summary() value)
# $STRUCTURE: object structure (str() value)
# $WARNING: warning messages (only if the warn.print argument is FALSE)
# If data is made of numerics, provide also:
# $INF.NB: number of Inf and -Inf
# $RANGE: range after removing Inf and NA
# $SUM: sum after removing Inf and NA
# $MEAN: mean after removing Inf and NA
# If data is a 2D object, provide also:
# $ROW_NAMES: row names
# $COL_NAMES: column names
# If data is a data frame, provide also:
# $COLUMN_TYPE: type of each column (typeof() value)
# If data is a list, provide also:
# $COMPARTMENT_NAMES: names of the comprtments
# $COMPARTMENT_TYPE: type of each compartment (typeof() value)
# REQUIRED PACKAGES
# None
# REQUIRED FUNCTIONS FROM THE cute PACKAGE
# fun_check()
# fun_get_message()
# EXAMPLE
# fun_info(data = 1:3)
# see http
# DEBUGGING
# mat1 <- matrix(1:3) ; data = env1 ; n = NULL ; warn.print = TRUE # for function debugging
# function name
function.name <- paste0(as.list(match.call(expand.dots = FALSE))[[1]], "()")
arg.names <- names(formals(fun = sys.function(sys.parent(n = 2)))) # names of all the arguments
arg.user.setting <- as.list(match.call(expand.dots = FALSE))[-1] # list of the argument settings (excluding default values not provided by the user)
# end function name
# required function checking
req.function <- c(
"fun_check"
)
tempo <- NULL
for(i1 in req.function){
if(length(find(i1, mode = "function")) == 0){
tempo <- c(tempo, i1)
}
}
if( ! is.null(tempo)){
tempo.cat <- paste0("ERROR IN ", function.name, "\nREQUIRED cute FUNCTION", ifelse(length(tempo) > 1, "S ARE", " IS"), " MISSING IN THE R ENVIRONMENT:\n", paste0(tempo, collapse = "()\n"))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end required function checking
# reserved words
# end reserved words
# arg with no default values
mandat.args <- c(
"data"
)
tempo <- eval(parse(text = paste0("missing(", paste0(mandat.args, collapse = ") | missing("), ")")))
if(any(tempo)){ # normally no NA for missing() output
tempo.cat <- paste0("ERROR IN ", function.name, "\nFOLLOWING ARGUMENT", ifelse(length(mandat.args) > 1, "S HAVE", "HAS"), " NO DEFAULT VALUE AND REQUIRE ONE:\n", paste0(mandat.args, collapse = "\n"))
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end arg with no default values
# argument primary checking
arg.check <- NULL #
text.check <- NULL #
checked.arg.names <- NULL # for function debbuging: used by r_debugging_tools
ee <- expression(arg.check <- c(arg.check, tempo$problem) , text.check <- c(text.check, tempo$text) , checked.arg.names <- c(checked.arg.names, tempo$object.name))
if( ! is.null(n)){
tempo <- fun_check(data = n, class = "vector", typeof = "integer", length = 1, double.as.integer.allowed = TRUE, fun.name = function.name) ; eval(ee)
}else{
# no fun_check test here, it is just for checked.arg.names
tempo <- fun_check(data = n, class = "vector")
checked.arg.names <- c(checked.arg.names, tempo$object.name)
}
tempo <- fun_check(data = warn.print, class = "logical", length = 1, fun.name = function.name) ; eval(ee)
if(any(arg.check) == TRUE){ # normally no NA
stop(paste0("\n\n================\n\n", paste(text.check[arg.check], collapse = "\n"), "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between == #
}
# source("C:/Users/Gael/Documents/Git_versions_to_use/debugging_tools_for_r_dev-v1.7/r_debugging_tools-v1.7.R") ; eval(parse(text = str_basic_arg_check_dev)) ; eval(parse(text = str_arg_check_with_fun_check_dev)) # activate this line and use the function (with no arguments left as NULL) to check arguments status and if they have been checked using fun_check()
# end argument primary checking
# second round of checking and data preparation
# management of NA arguments
tempo.arg <- names(arg.user.setting) # values provided by the user
tempo.log <- suppressWarnings(sapply(lapply(lapply(tempo.arg, FUN = get, env = sys.nframe(), inherit = FALSE), FUN = is.na), FUN = any)) & lapply(lapply(tempo.arg, FUN = get, env = sys.nframe(), inherit = FALSE), FUN = length) == 1 # no argument provided by the user can be just NA
if(any(tempo.log) == TRUE){ # normally no NA because is.na() used here
tempo.cat <- paste0("ERROR IN ", function.name, ":\n", ifelse(sum(tempo.log, na.rm = TRUE) > 1, "THESE ARGUMENTS\n", "THIS ARGUMENT\n"), paste0(tempo.arg[tempo.log], collapse = "\n"),"\nCANNOT JUST BE NA")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end management of NA arguments
# management of NULL arguments
tempo.arg <-c(
"data", 
# "n", # because can be NULL
"warn.print"
)
tempo.log <- sapply(lapply(tempo.arg, FUN = get, env = sys.nframe(), inherit = FALSE), FUN = is.null)
if(any(tempo.log) == TRUE){# normally no NA with is.null()
tempo.cat <- paste0("ERROR IN ", function.name, ":\n", ifelse(sum(tempo.log, na.rm = TRUE) > 1, "THESE ARGUMENTS\n", "THIS ARGUMENT\n"), paste0(tempo.arg[tempo.log], collapse = "\n"),"\nCANNOT BE NULL")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}
# end management of NULL arguments
# code that protects set.seed() in the global environment
# end code that protects set.seed() in the global environment
# warning initiation
ini.warning.length <- options()$warning.length
options(warning.length = 8170)
warn <- NULL
# warn.count <- 0 # not required
# end warning initiation
# other checkings
if( ! is.null(n)){
if(n < 1){
tempo.cat <- paste0("ERROR IN ", function.name, ": n ARGUMENT MUST BE A POSITIVE AND NON NULL INTEGER")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}else if(is.finite(n)){
# warn.count <- warn.count + 1
tempo.warn <- paste0("SOME COMPARTMENTS CAN BE TRUNCATED (n ARGUMENT IS ", n, ")")
warn <- paste0(ifelse(is.null(warn), tempo.warn, paste0(warn, "\n\n", tempo.warn)))
}
}
# end other checkings
# reserved word checking
# end reserved word checking
# end second round of checking and data preparation
# package checking
# end package checking
# main code
# new environment
env.name <- paste0("env", as.numeric(Sys.time()))
if(exists(env.name, where = -1)){ # verify if still ok when fun_info() is inside a function
tempo.cat <- paste0("ERROR IN ", function.name, ": ENVIRONMENT env.name ALREADY EXISTS. PLEASE RERUN ONCE")
stop(paste0("\n\n================\n\n", tempo.cat, "\n\n================\n\n"), call. = FALSE) # == in stop() to be able to add several messages between ==
}else{
assign(env.name, new.env())
assign("data", data, envir = get(env.name, env = sys.nframe(), inherit = FALSE)) # data assigned in a new envir for test
}
# end new environment
data.name <- deparse(substitute(data))
output <- list("NAME" = data.name)
tempo.try.error <- fun_get_message(data = "class(data)", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
tempo <- list("CLASS" = class(data))
output <- c(output, tempo)
}
tempo.try.error <- fun_get_message(data = "typeof(data)", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
tempo <- list("TYPE" = typeof(data))
output <- c(output, tempo)
}
tempo.try.error <- fun_get_message(data = "length(data)", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
tempo <- list("LENGTH" = length(data))
output <- c(output, tempo)
}
if(all(typeof(data) %in% c("integer", "numeric", "double")) & ! any(class(data) %in% "factor")){ # all() without na.rm -> ok because typeof(NA) is "logical" # any() without na.rm -> ok because class(NA) is "logical"
tempo <- list("INF.NB" = sum(is.infinite(data)))
output <- c(output, tempo)
tempo <- list("RANGE" = range(data[ ! is.infinite(data)], na.rm = TRUE))
output <- c(output, tempo)
tempo <- list("SUM" = sum(data[ ! is.infinite(data)], na.rm = TRUE))
output <- c(output, tempo)
tempo <- list("MEAN" = mean(data[ ! is.infinite(data)], na.rm = TRUE))
output <- c(output, tempo)
}
if(all(typeof(data) %in% c("logical", "integer", "double", "complex", "character", "list"))){ # all() without na.rm -> ok because typeof(NA) is "logical"
tempo.try.error <- fun_get_message(data = "is.na(data)", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
tempo <- list("NA.NB" = sum(is.na(data)))
output <- c(output, tempo)
}
}
tempo.try.error <- fun_get_message(data = "head(data)", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
tempo <- list("HEAD" = head(data))
output <- c(output, tempo)
tempo <- list("TAIL" = tail(data)) # no reason that tail() does not work if head() works
output <- c(output, tempo)
}
tempo.try.error <- fun_get_message(data = "dim(data)", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
if(length(dim(data)) > 0){
tempo <- list("DIMENSION" = dim(data))
if(length(tempo[[1]]) == 2){
names(tempo[[1]]) <- c("NROW", "NCOL")
}
output <- c(output, tempo)
}
}
if(all(class(data) == "data.frame") | all(class(data) %in% c("matrix", "array")) | all(class(data) == "table")){ # all() without na.rm -> ok because typeof(NA) is "logical"
if(length(dim(data)) > 1){ # to avoid 1D table
tempo <- list("ROW_NAMES" = dimnames(data)[[1]])
output <- c(output, tempo)
tempo <- list("COLUM_NAMES" = dimnames(data)[[2]])
output <- c(output, tempo)
}
}
tempo.try.error <- fun_get_message(data = "summary(data)", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
tempo <- list("SUMMARY" = summary(data))
output <- c(output, tempo)
}
tempo.try.error <- fun_get_message(data = "noquote(matrix(capture.output(str(data))))", kind = "error", header = FALSE, env = get(env.name, env = sys.nframe(), inherit = FALSE))
if(is.null(tempo.try.error)){
tempo <- capture.output(str(data))
tempo <- list("STRUCTURE" = noquote(matrix(tempo, dimnames = list(rep("", length(tempo)), "")))) # str() print automatically, ls.str() not but does not give the order of the data.frame
output <- c(output, tempo)
}
if(all(class(data) == "data.frame")){ # all() without na.rm -> ok because class(NA) is "logical"
tempo <- list("COLUMN_TYPE" = sapply(data, FUN = "typeof"))
if(any(sapply(data, FUN = "class") %in% "factor")){ # if an ordered factor is present, then sapply(data, FUN = "class") return a list but works with any(sapply(data, FUN = "class") %in% "factor") # any() without na.rm -> ok because class(NA) is "logical"
tempo.class <- sapply(data, FUN = "class")
if(any(unlist(tempo.class) %in% "ordered")){ # any() without na.rm -> ok because class(NA) is "logical"
tempo2 <- sapply(tempo.class, paste, collapse = " ") # paste the "ordered" factor" in "ordered factor"
}else{
tempo2 <- unlist(tempo.class)
}
tempo[["COLUMN_TYPE"]][grepl(x = tempo2, pattern = "factor")] <- tempo2[grepl(x = tempo2, pattern = "factor")]
}
output <- c(output, tempo)
}
if(all(class(data) == "list")){ # all() without na.rm -> ok because class(NA) is "logical"
tempo <- list("COMPARTMENT_NAMES" = names(data))
output <- c(output, tempo)
tempo <- list("COMPARTMENT_TYPE" = sapply(data, FUN = "typeof"))
if(any(unlist(sapply(data, FUN = "class")) %in% "factor")){ # if an ordered factor is present, then sapply(data, FUN = "class") return a list but works with any(sapply(data, FUN = "class") %in% "factor")  # any() without na.rm -> ok because class(NA) is "logical"
tempo.class <- sapply(data, FUN = "class")
if(any(unlist(tempo.class) %in% "ordered")){ # any() without na.rm -> ok because class(NA) is "logical"
tempo2 <- sapply(tempo.class, paste, collapse = " ") # paste the "ordered" factor" in "ordered factor"
}else{
tempo2 <- unlist(tempo.class)
}
tempo[["COMPARTMENT_TYPE"]][grepl(x = tempo2, pattern = "factor")] <- tempo2[grepl(x = tempo2, pattern = "factor")]
}
output <- c(output, tempo)
}
if( ! is.null(n)){
output[names(output) != "STRUCTURE"] <- lapply(X = output[names(output) != "STRUCTURE"], FUN = head, n = n, simplify = FALSE)
}
# output
if(warn.print == FALSE){
output <- c(output, WARNING = warn)
}else if(warn.print == TRUE & ! is.null(warn)){
on.exit(warning(paste0("FROM ", function.name, ":\n\n", warn), call. = FALSE))
}
on.exit(exp = options(warning.length = ini.warning.length), add = TRUE)
return(output)
# end output
# end main code
}


######## fun_head() #### head of the left or right of big 2D objects


fun_head <- function(
data1, 
n = 6, 
side = "l"
){
# AIM
# as head() but display the left or right head of big 2D objects
# ARGUMENTS
# data1: any object but more dedicated for matrix, data frame or table
# n: as in head() but for for matrix, data frame or table, number of dimension to print (10 means 10 rows and columns)
# side: either "l" or "r" for the left or right side of the 2D object (only for matrix, data frame or table)
# BEWARE: other arguments of head() not used
# RETURN
# the head
# REQUIRED PACKAGES
# none
# REQUIRED FUNCTIONS FROM CUTE_LITTLE_R_FUNCTION
# fun_check()
# EXAMPLES
# obs1 = matrix(1:30, ncol = 5, dimnames = list(letters[1:6], LETTERS[1:5])) ; obs1 ; fun_head(obs1, 3)
# obs1 = matrix(1:30, ncol = 5, dimnames = list(letters[1:6], LETTERS[1:5])) ; obs1 ; fun_head(obs1, 3, "right")
# DEBUGGING
# data1 = matrix(1:30, ncol = 5) # for function debugging
# data1 = matrix(1:30, ncol = 5, dimnames = list(letters[1:2], LETTERS[1:5])) # for function debugging