R/input_RVP.R
In FabienNicol/suppliesdash: EFS LFP Sales Dashboard
# --- PLOTTING PIPELINE FUNCTIONS ---------------------------------------------
#' Remove, or not, the size column.
#'
#' \code{f_keepSize} is used in \code{\link{data4plot}}.
#'
#' @param keepSize A boolean
#' @param data A dataframe
#' @return a dataframe.
f_keepSize <- function(data, keepSize) {
if (isTRUE(keepSize)) data else data %>% dplyr::select(-size)
}
#' Remove, or not, the column source.
#'
#' \code{f_keepSource} is used in \code{\link{data4plot}}.
#' In \code{f_keepSource}, if the source column is kept, it is mutated into an
#' ordered factor.
#'
#' @param keepSource A boolean
#' @param data A dataframe
#' @return a dataframe.
f_keepSource <- function(data, keepSource) {
if (isTRUE(keepSource)) {
data %>%
dplyr::mutate(source = factor(source, c("Transac", "QT", "MPS", "Store")))
} else {
data %>%
dplyr::select(-source)
}
}
#' Keep the relevant region columns (1 or 2 distinct).
#'
#' \code{f_keepRegion} is used in \code{\link{data4plot}}.
#' If ReportRegionLevel == RegionSplitLevel, the 2 output region colums will
#' have the same value.
#' NA values and "PEA" are removed silently.
#'
#' @param ReportRegionLevel A string, such as "RegionLevel0".
#' @param ReportRegionValue A string, such as "EMEAR".
#' @param RegionSplitLevel A string, such as "RegionLevel2".
#' @param data A dataframe with columns RegionLevel0, 1, 2, 4.
#' @return a dataframe with columns RegionReport, RegionSplit.
f_keepRegion <- function(data, ReportRegionLevel, ReportRegionValue, RegionSplitLevel) {
regionFacet <- (ReportRegionLevel == RegionSplitLevel)
ReportRegionLevel <- dplyr::sym(ReportRegionLevel)
RegionSplitLevel <- dplyr::sym(RegionSplitLevel)
data %>%
dplyr::mutate(RegionSplit = !!RegionSplitLevel) %>%
dplyr::mutate(temp = !!ReportRegionLevel) %>%
# take into account the case where RRL == RSL (no facet on region)
dplyr::mutate(RegionReport = if (regionFacet) RegionSplit else temp) %>%
dplyr::filter(RegionReport == ReportRegionValue,
RegionSplit != "PEA",
!is.na(RegionSplit)) %>%
dplyr::select_at(dplyr::vars(-dplyr::matches("^RegionLevel|temp")))
}
#' Keep the relevant supplies columns (1 or 2 distinct).
#'
#' \code{f_keepSupplies} is used in \code{\link{data4plot}}.
#' If SuppliesPlotLevel == SuppliesSplitLevel, the 2 output supplies columns
#' are identical.
#' NA values are removed silently.
#'
#' @param data a dataframe with columns SuppliesLevel0, 1, 2, 3, 4.
#' @param SuppliesPlotLevel a string, such as "SuppliesLevel0".
#' @param SuppliesPlotValue a string, such as "LFP aqueous ink".
#' @param SuppliesSplitLevel a string, such as "SuppliesLevel3".
#' @return a dataframe with columns SuppliesPlot, SuppliesSplit.
f_keepSupplies <- function(data, SuppliesPlotLevel, SuppliesPlotValue, SuppliesSplitLevel) {
SuppliesPlotLevel <- dplyr::sym(SuppliesPlotLevel)
SuppliesSplitLevel <- dplyr::sym(SuppliesSplitLevel)
data %>%
dplyr::mutate(SuppliesSplit = !!SuppliesSplitLevel,
SuppliesPlot = !!SuppliesPlotLevel) %>%
dplyr::filter(!is.na(SuppliesSplit),
SuppliesPlot == SuppliesPlotValue) %>%
dplyr::select_at(dplyr::vars(-dplyr::matches("^SuppliesLevel")))
}
#' Rank Region according to revenue (or volume) within the reference year.
#'
#' \code{f_rkRegion} is used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns: fDate, RegionSplit, Rev, Vol.
#' @param yearRef an Integer representing the reference year.
#' @param rankBy a string representing the measure on which the Region are
#' ranked, "Rev" or "Vol".
#' @return a dataframe with an ordered factor for column RegionSplit and
#' column fYear.
f_rkRegion <- function(data, yearRef = 2017, rankBy = "Rev") {
rankBy <- dplyr::sym(rankBy)
data2 <-
data %>%
dplyr::mutate(fYear = lubridate::year(fDate)) %>%
dplyr::group_by(RegionSplit) %>%
dplyr::mutate(Region_rk = sum(ifelse(fYear == yearRef, (!!rankBy), 0), na.rm = T)) %>%
dplyr::ungroup()
# pipeline fail for intance for regionreportvalue=="EFS" and inkreportvalue=="G&J", need to break here
if (nrow(data2) == 0) return("no data f_rkRegion")
data2 %>%
dplyr::mutate(RegionSplit = forcats::fct_reorder(RegionSplit, Region_rk, .desc = TRUE)) %>%
dplyr::select(-Region_rk)
}
#' Rank supplies according to revenue (or volume) within the reference year.
#'
#' \code{f_rkSupplies} is used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns: fYear, SuppliesSplit, SuppliesPlot,
#' Rev, Vol.
#' @param yearRef an Integer representing the reference year.
#' @param rankBy a string representing the measure on which the Region are
#' ranked, "Rev" or "Vol".
#' @return a dataframe with an ordered factor for column SuppliesSplit
#' and fYear.
f_rkSupplies <- function(data, yearRef = 2017, rankBy = "Rev") {
# the case of "no data ..."
if (is.character(data)) return(data)
rankBy <- dplyr::sym(rankBy)
data2 <-
data %>%
dplyr::group_by(SuppliesPlot, SuppliesSplit) %>%
dplyr::mutate(supplies_rk = sum(ifelse(fYear == yearRef, (!!rankBy), 0), na.rm = T)) %>%
dplyr::ungroup()
if (nrow(data2) == 0) return("no data f_rkSupplies")
data2 %>%
dplyr::mutate(rk = dplyr::dense_rank(dplyr::desc(supplies_rk)),
SuppliesSplit = paste0(rk, "-", SuppliesSplit),
SuppliesSplit = forcats::fct_reorder(
SuppliesSplit, supplies_rk, .desc = TRUE)
) %>%
dplyr::select(-supplies_rk)
}
#' Keep the topN supplies items and lump together the Others.
#'
#' \code{f_topSupplies} keeps the topN items on SuppliesSplit column.
#' Used in \code{\link{data4plot}}.
#'
#' @param data A dataframe with columns rk, SuppliesSplit.
#' @param topN An integer representing the number of items to display, the
#' others will be grouped together.
#' @return a dataframe with a column SuppliesSplit, an ordered factor.
f_topSupplies <- function(data, topN = 10) {
# the case of "no data ..."
if (is.character(data)) return(data)
#above, return(data %>% dplyr::select(-rk)), will generate:
#no applicable method for 'select_' applied to an object of class "character"
if (is.null(topN)) return(data)
data %>%
{dplyr::bind_rows(dplyr::filter(., rk <= topN),
dplyr::filter(., rk > topN) %>%
dplyr::mutate(SuppliesSplit = "Others", rk = topN+1)
# NO NEED TO summarise (it's done later)
# %>% group_by(SuppliesSplit, fYear, RegionSplit, rk) %>%
# summarise_if(is.numeric, sum, na.rm = T) %>%
# ungroup
) %>%
dplyr::mutate(SuppliesSplit = forcats::fct_reorder(
SuppliesSplit, rk, .desc = F)) %>%
dplyr::select(-rk)
# ,label = paste0(round(Liters / roundMultiple, ifelse(abs(Liters) < roundThreshold, 1, 0))))
}
}
#' Removes a column to a dataframe.
#'
#' \code{f_keepDateYear} removes a column to a dataframe.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe
#' @param ditch a string, a column name.
#' @return a dataframe without the column ditch.
f_keepDateYear <- function(data, ditch = "fYear") {
# the case of "no data ..."
if (is.character(data)) return(data)
ditch <- dplyr::sym(ditch)
data %>% dplyr::select(-(!!ditch))
}
#' Summarise all numeric columns (sum)
#'
#' \code{f_summariseIFnum} is grouping by the non numeric columns and
#' summarises, with a sum, the numeric columns. If there is a column fYear
#' the column is converted in text during the calculation and return as an
#' integer.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe
#' @return a dataframe
f_summariseIFnum <- function(data) {
# the case of "no data ..."
if (is.character(data)) return(data)
# mutate fYear as character for column selection (group_by_if)
if ("fYear" %in% names(data)) {
data <- data %>% dplyr::mutate(fYear = as.character(fYear))
}
data2 <-
data %>%
# numeric can be Rev, Rev_sm, Liters, Liters_sm
dplyr::group_by_if(~ !is.numeric(.x)) %>%
dplyr::summarise_if(is.numeric, sum, na.rm = T) %>%
dplyr::ungroup()
if ("fYear" %in% names(data)) {
data2 <- data2 %>% dplyr::mutate(fYear = as.integer(fYear))
}
return(data2)
}
#' Prepare data for lollipop plots.
#'
#' \code{f_summarise4lollipop} adds information at level (RegionSplit, fYear),
#' and level (RegionSplit, SuppliesSplit, fYear), while keeping all the rows of
#' the data frame.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with column fYear, lollipop_measure, regionSplit,
#' SuppliesSplit
#' @param lollipop_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a lollipop.
#' @param showFY a string vector, such as "All" or c("2017", "2018").
#' @return a dataframe with columns RegionSplit, SuppliesSplit, fYear,
#' percentage, measure.
f_summarise4lollipop <- function(data, lollipop_measure = "Rev", showFY) {
# lollipop_measure: Null if not lollipop chart, Rev if based on revenue, Liters if based on volume (Liters, sqm)
# the case of "no data ..."
if (is.character(data)) return(data)
if (is.null(lollipop_measure)) return(data)
lollipop_measure <- dplyr::sym(lollipop_measure)
data %>%
dplyr::filter(if (showFY != "All") fYear %in% showFY else TRUE) %>%
dplyr::mutate(measure = !!lollipop_measure) %>%
dplyr::group_by(RegionSplit, fYear) %>%
dplyr::mutate(tot = sum(measure, na.rm = T)) %>%
dplyr::ungroup() %>%
dplyr::group_by(RegionSplit, SuppliesSplit, fYear) %>%
dplyr::mutate(percentage = sum(measure, na.rm = T) / tot) %>%
dplyr::ungroup() %>%
# remove tot, x_sm (and the one not selected among RevVol)
dplyr::select_at(dplyr::vars(-dplyr::matches("^tot$|_sm$")))
}
#' Prepare data for time series plots.
#'
#' \code{f_summarise4ts} filters out old data ans a few old models.
#' Identify the column measure.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns fDate, ts_measure, SuppliesSplit.
#' @param ts_measure a string, such as "Rev_sm" or "Liters_sm". NULL if not a
#' ts plot.
#' @return a dataframe with column fDate, measure, SuppliesSplit.
f_summarise4ts <- function(data, ts_measure = "Rev_sm") {
# the case of "no data ..."
if (is.character(data)) return(data)
if (is.null(ts_measure)) return(data)
# remove the (window-1)th first points of the smoothed values
date2 <- "2014/05/03"
ts_measure <- dplyr::sym(ts_measure)
data2 <-
data %>%
dplyr::filter(# DELETE fDate < lubridate::ymd(date1),
fDate > lubridate::ymd(date2),
!(stringr::str_detect(SuppliesSplit,
"T549_10600|T46X_7000|T47X_9500|T480_7500|T486_5500|T500_10000|T40X_9000"))) %>%
dplyr::mutate(measure = !!ts_measure)
if (nrow(data2) == 0) return("no data f_summarise4ts")
return(data2)
# MUST BE DONE ALREADY IN f_rkSupplies
# group_by(SuppliesPlot, SuppliesSplit, fYear, RegionReport, RegionSplit, fDate) %>%
# summarise(Liters = sum(Liters.sm, na.rm = T),
# LitersFY = sum(Liters, na.rm = T)) %>%
# ungroup %>%
# group_by(SuppliesPlot, SuppliesSplit) %>%
# mutate(ink_rk = sum(Liters, na.rm = T)) %>%
# ungroup
# if (nrow(IS1) == 0) return("no data")
#
# IS1 <- IS1 %>%
# mutate(rk = dense_rank(desc(ink_rk)),
# SuppliesSplit = paste0(rk, "-", SuppliesSplit),
# SuppliesSplit = forcats::fct_reorder(SuppliesSplit, ink_rk, .desc = TRUE)
# # InkLevel = paste0(as.character(as.numeric(InkLevel)), "-", InkLevel)
# ) %>%
# select(-ink_rk, rk)
# ADD THOSE INFORMATION LATER
# find region_RHS, place a label with the inkLevel name at the end of the line of the most RHS region
# also compute the % per region, to be included in the strip.text of the facet
# IS2 <- IS1 %>%
# filter(fYear == yearRef) %>%
# group_by(RegionSplit) %>%
# summarise(LitersFY = sum(LitersFY, na.rm = T)) %>%
# ungroup %>%
# arrange(desc(LitersFY)) %>%
# mutate(perc = 100 * LitersFY/sum(LitersFY))
# region_RHS <- IS2 %>%
# #first, wrapper around [[
# select(RegionSplit) %>% unlist %>% last
#
# IS1 <- IS1 %>%
# group_by(RegionSplit) %>%
# mutate(label = if_else(fDate == max(fDate) & RegionSplit == region_RHS, as.character(as.numeric(SuppliesSplit)), NA_character_))
#
# add in the strip.text of the facet the % represented by the region
# if (length(levels(IS1$RegionSplit)) != nrow(IS2)) print("inkLiLj. Error with number of levels in region_RHS and yearRef")
# if (length(levels(IS1$RegionSplit))>1) levels(IS1$RegionSplit) <- paste0(levels(IS1$RegionSplit), "\n", round(IS2$perc,0), "% in FY", yearRef)
}
#' Prepare data for mosaic plots.
#'
#' \code{f_summarise4mosaic} filters the fiscal years, identifies the measure
#' column.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns source, RegionSplit, RegionReport,
#' SuppliesSplit, SuppliesPlot, fYear, mosaic_measure.
#' @param mosaic_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a mosaic.
#' @param showFY a string vector, such as "All" or c("2017", "2018")
#' @return a dataframe with columns source, RegionSplit, RegionReport,
#' SuppliesSplit, SuppliesPlot, fYear, measure.
f_summarise4mosaic <- function(data, mosaic_measure, showFY) {
# the case of "no data ..."
if (is.character(data)) return(data)
if (is.null(mosaic_measure)) return(data)
mosaic_measure <- dplyr::sym(mosaic_measure)
data %>%
dplyr::mutate(measure = !!mosaic_measure) %>%
dplyr::filter(if (showFY != "All") fYear %in% showFY else TRUE) %>%
dplyr::select(source, RegionSplit, RegionReport, SuppliesSplit, SuppliesPlot, fYear, measure)
}
#' Prepare data for RVP plots.
#'
#' \code{f_summarise4RVP} adds information at level (RegionSplit, fYear),
#' and level (RegionSplit, SuppliesSplit, fYear), while keeping all the rows of
#' the data frame.
#' Requires to drop fYear before (\code{f_keepDateYear}).
#'
#' @param data a dataframe with columns SuppliesSplit, RegionReport, Supplies,
#' facetY, Rev, Liters, Rev_sm, Liters_sm, Date, fDate.
#' @param facetY a string, such as "RegionSplit", "source". NULL is the plot
#' isn't an RVP.
#' @return a dataframe with columns SuppliesSplit, RegionReport, Supplies,
#' facetY, KEY, VALUE, Date, fDate, label. KEY (Rev_sm, Liters_sm, price) a
#' factor.
f_summarise4RVP <- function(data, facetY) {
if (is.null(facetY)) return(data)
strip_sep <- ifelse(facetY == "size", "\n", "*")
# cat(file = stderr(), "f_summarise4RVP > facetY: ", facetY, ", selected sep: ", strip_sep, "\n")
# remove the first (window-1)th month
date2 <- "2014/05/03"
data %>%
dplyr::filter(fDate > lubridate::ymd(date2),
!(stringr::str_detect(SuppliesSplit,
"T549_10600|T46X_7000|T47X_9500|T480_7500|T486_5500|T500_10000|T40X_9000"))) %>%
dplyr::mutate(facetY = !!dplyr::sym(facetY)) %>%
dplyr::select(-Rev, -Liters) %>%
dplyr::group_by_at(dplyr::vars(dplyr::matches(
"facetY|RegionReport|Supplies|Rev_sm|Liters_sm|Date"))) %>%
dplyr::summarise_if(is.numeric, dplyr::funs(sum(., na.rm = T))) %>%
dplyr::ungroup() %>%
dplyr::mutate(price = ifelse(Liters_sm == 0 | is.na(Liters_sm), 0, Rev_sm / Liters_sm)) %>%
tidyr::gather(key = KEY, value = VALUE, Rev_sm, Liters_sm, price) %>%
dplyr::mutate(KEY = factor(KEY, c("Rev_sm", "Liters_sm", "price"))) %>%
# size = factor(size, c("RA4", "R13", "R24", "R36", "R44", "R64", "SA4", "SA3", "SA2", "SA1"))
#factor created in the order of appearance wit forcarts::as_factor
dplyr::arrange(KEY, facetY) %>%
dplyr::mutate(label = paste0(KEY, strip_sep, facetY),
label = forcats::as_factor(label))
}
# ===< test ===
# dat <- read_csv2("./temp/inkSalesEnriched_RVP.csv", col_types = "ccDccccccccccdddd")
# dat2 <- dat %>%
# f_keepSize(TRUE) %>%
# f_keepSource(FALSE) %>%
# f_keepRegion(ReportRegionLevel = "RegionLevel2", ReportRegionValue = "EU5-NB", RegionSplitLevel = "RegionLevel4") %>%
# f_keepSupplies(SuppliesPlotLevel = "SuppliesLevel1", SuppliesPlotValue = "LFPmedia", SuppliesSplitLevel = "SuppliesLevel2") %>%
# f_rkRegion(yearRef = 2017, rankBy = "Rev") %>%
# f_rkSupplies(yearRef = 2017, rankBy = "Rev") %>%
# f_topSupplies(topN = 10) %>%
# f_keepDateYear(ditch = "fYear") %>%
# f_summariseIFnum %>%
# f_summarise4lollipop(lollipop_measure = NULL, showFY = "All") %>%
# f_summarise4ts(ts_measure = NULL) %>%
# f_summarise4mosaic(mosaic_measure = NULL, showFY = "All") %>%
# f_summarise4RVP(facetY = "size")
# -----
# lollipop_input <- data4plot_Lollipop(dat, RRL = "RegionLevel2", RRV = "EU5-NB", RSL = "RegionLevel2",
# SPL = "SuppliesLevel1", SPV = "aqueous", SSL = "SuppliesLevel2",
# yearRef = 2017, rankBy = "Rev", topN = 10,
# lollipop_measure = "Rev")
# p <- plot_lollipop(lollipop_input$data, lollipop_input$labs)
# -----
# ts_input <- data4plot_ts(dat, RRL="RegionLevel2", RRV="EU5-NB", RSL="RegionLevel2",
# SPL="SuppliesLevel1", SPV="aqueous", SSL="SuppliesLevel2",
# yearRef=2017, rankBy="Rev", topN=10, ts_measure="Rev")
# plot_ts(ts_input$data, ts_input$labs)
# -----
# mosaic_input <- data4plot_Mosaic(dat, RRL="RegionLevel2", RRV="EU5-NB", RSL="RegionLevel2",
# SPL="SuppliesLevel1", SPV="aqueous", SSL="SuppliesLevel2",
# yearRef=2017, showFY=2017, rankBy="Rev", topN=5, mosaic_measure="Rev")
# plot_mosaic(mosaic_input$data, mosaic_input$labs)
# plot_mosaic(dat2, c("labs1", "labs2", "labs3")) %>% ggplotly
# === end test >===
#' Prepare data for plot lollipop, ts, mosaic, RVP.
#'
#' \code{data4plot} regroups several functions, in order:
#' \code{\link{f_keepSize}}, \code{\link{f_keepSource}},
#' \code{\link{f_keepRegion}}, \code{\link{f_keepSupplies}},
#' \code{\link{f_rkRegion}}, \code{\link{f_rkSupplies}},
#' \code{\link{f_topSupplies}}, \code{\link{f_keepDateYear}},
#' \code{\link{f_summariseIFnum}}, \code{\link{f_summarise4lollipop}},
#' \code{\link{f_summarise4ts}}, \code{\link{f_summarise4mosaic}},
#' \code{\link{f_summarise4RVP}}.
#' \code{data4plot} is used in:
#' \code{\link{data4plot_RVP}}
#' \code{\link{data4plot_Lollipop}}
#' \code{\link{data4plot_Mosaic}}
#' \code{\link{data4plot_ts}}
#'
#' @param data a dataframe.
#' @param keepSize a boolean.
#' @param keepSource a boolean.
#' @param RRL a string, reportRegionLevel.
#' @param RRV a string, ReportRegionValue.
#' @param RSL a string, RegionSplitLevel.
#' @param SPL a string, SuppliesPlotLevel.
#' @param SPV a string, SuppliesPlotValue.
#' @param SSL a string, SppliesSplitLevel.
#' @param yearRef an Integer representing the reference year.
#' @param rankBy a string representing the measure on which the Region are
#' ranked, "Rev" or "Vol".
#' @param topN An integer representing the number of items to display, the
#' others will be grouped together.
#' @param ditch a string, a column name.
#' @param showFY a string vector, such as "All" or c("2017", "2018").
#' @param lollipop_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a lollipop.
#' @param ts_measure a string, such as "Rev_sm" or "Liters_sm". NULL if not a
#' ts plot.
#' @param mosaic_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a mosaic.
#' @param RVP_facetY a string, such as "RegionSplit", "source". NULL is the
#' plot isn't an RVP.
#' @return a dataframe expected by one of the plot functions.
data4plot <- function(data, keepSize, keepSource,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch, showFY,
lollipop_measure, ts_measure, mosaic_measure, RVP_facetY) {
# DEBUG
# cat(file = stderr(), "data4plot > RP_facetY: ", RVP_facetY, "\n")
data2 <- data %>%
f_keepSize(keepSize) %>%
f_keepSource(keepSource) %>%
f_keepRegion(ReportRegionLevel = RRL,
ReportRegionValue = RRV,
RegionSplitLevel = RSL) %>%
f_keepSupplies(SuppliesPlotLevel = SPL,
SuppliesPlotValue = SPV,
SuppliesSplitLevel = SSL) %>%
f_rkRegion(yearRef, rankBy) %>%
f_rkSupplies(yearRef, rankBy) %>%
f_topSupplies(topN) %>%
f_keepDateYear(ditch) %>%
f_summariseIFnum %>%
f_summarise4lollipop(lollipop_measure, showFY) %>%
f_summarise4ts(ts_measure) %>%
f_summarise4mosaic(mosaic_measure, showFY) %>%
f_summarise4RVP(RVP_facetY)
return(data2)
}
#' Returns data for plot_RVP.
#'
#' \code{data4plot_RVP} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @param free_y a string, such as "free_y".
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_RVP <- function(data, keepSize, keepSource, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN,
RVP_facetY,
SP_filter_part, free_y) {
data2 <- data4plot(data, keepSize, keepSource,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fYear", showFY = "All",
lollipop_measure = NULL, ts_measure = NULL,
mosaic_measure = NULL, RVP_facetY)
labs <- list(RVP_facetY = RVP_facetY, SPV = SPV, SP_filter_part = SP_filter_part, free_y = free_y)
return(list(data = data2, labs = labs))
}
#' Returns the order or magnitude of a double (1, 1 000, 100 000, ...).
#'
#' \code{divid} is used \code{\link{data4plot_Lollipop}}.
#'
#' @param y a double.
#' @return an integer representing the order of magnitude.
divid <- function(y) {
signif(y,3) %>% log10 %>% `%/%`(3) %>% floor %>% (function(x) {10^(3*x)})
}
#' Returns data for plot_lollipop.
#'
#' \code{data4plot_Lollipop} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_Lollipop <- function(data, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, showFY,
lollipop_measure,
SP_filter_part) {
data2 <- data4plot(data, keepSize = F, keepSource = F,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fDate", showFY,
lollipop_measure, ts_measure = NULL,
mosaic_measure = NULL, RVP_facetY = NULL)
div <- data2 %>%
dplyr::select(measure) %>% unlist %>% max(na.rm = T) %>% divid
vol_unit <- data_SL2_regex[data_SL2_regex$SuppliesLevel2_type == SP_filter_part, "unit"][[1]]
measure_txt <- ifelse(stringr::str_detect(lollipop_measure, "Rev"), "Revenue", "Volume")
vol_unit <- ifelse(measure_txt == "Revenue", "Euros NetNet", vol_unit)
labs <- list(title = paste0(measure_txt,
" (x ", format(div, scientific = FALSE, big.mark = " "), " ", vol_unit,
") per FY. Rank according to FY", yearRef),
subtitle = paste0("sell-in data for the top ", topN, " models in product group ", SPV, "."),
x_lab = "% of total",
y_lab = "unit",
not_use1 = "Ink / Printer model",
caption = paste0(RRL, RRV, RSL,"_", SPL, SPV, SSL, "_", yearRef,
lollipop_measure, collapse = "_"),
div = div)
# labs <- c(paste0("Volume (x", div," sqm or Liters) per FY. Rank according to FY", yearRef),
# paste0("sell-in data for the top ", topN, " models in product group ", SPV, "."),
# "% of total ",
# "unit",
# "Ink / Printer model",
# paste0(RRL, RRV, RSL,"_", SPL, SPV, SSL, "_", yearRef,
# lollipop_measure, collapse = "_"),
# div)
return(list(data = data2, labs = labs))
}
#' Returns data for plot_mosaic.
#'
#' \code{data4plot_Mosaic} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_Mosaic <- function(data, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, showFY,
mosaic_measure,
SP_filter_part) {
data2 <- data4plot(data, keepSize = F, keepSource = T,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fDate", showFY,
lollipop_measure = NULL, ts_measure = NULL,
mosaic_measure, RVP_facetY = NULL)
vol_unit <- data_SL2_regex[data_SL2_regex$SuppliesLevel2_type == SP_filter_part, "unit"][[1]]
measure_txt <- ifelse(stringr::str_detect(mosaic_measure, "Rev"), "Revenue", "Volume")
vol_unit <- ifelse(measure_txt == "Revenue", "Euros NetNet", vol_unit)
labs <- c(paste0("Split of ", measure_txt, " (", vol_unit, "), per categories (within ", SPV, ") and channel ", RRV),
paste0("Top", topN, " categories | QTs are QTs approved by EMD"),
"plotNameMosaic")
#test
write.csv2(data2, "test_mosaic_jan18.csv")
return(list(data = data2, labs = labs))
}
#' Returns data for plot_ts.
#'
#' \code{data4plot_ts} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @param free_y a string, such as "free_y".
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_ts <- function(data, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN,
ts_measure,
SP_filter_part, free_y) {
data2 <- data4plot(data, keepSize = F, keepSource = F,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fYear", showFY = "All",
lollipop_measure = NULL, ts_measure,
mosaic_measure = NULL, RVP_facetY = NULL)
labs <- list(SPV = SPV, SP_filter_part = SP_filter_part, ts_measure = ts_measure, free_y = free_y)
return(list(data = data2, labs = labs))
}
FabienNicol/suppliesdash documentation built on May 6, 2019, 7:04 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
# --- PLOTTING PIPELINE FUNCTIONS ---------------------------------------------
#' Remove, or not, the size column.
#'
#' \code{f_keepSize} is used in \code{\link{data4plot}}.
#'
#' @param keepSize A boolean
#' @param data A dataframe
#' @return a dataframe.
f_keepSize <- function(data, keepSize) {
if (isTRUE(keepSize)) data else data %>% dplyr::select(-size)
}
#' Remove, or not, the column source.
#'
#' \code{f_keepSource} is used in \code{\link{data4plot}}.
#' In \code{f_keepSource}, if the source column is kept, it is mutated into an
#' ordered factor.
#'
#' @param keepSource A boolean
#' @param data A dataframe
#' @return a dataframe.
f_keepSource <- function(data, keepSource) {
if (isTRUE(keepSource)) {
data %>%
dplyr::mutate(source = factor(source, c("Transac", "QT", "MPS", "Store")))
} else {
data %>%
dplyr::select(-source)
}
}
#' Keep the relevant region columns (1 or 2 distinct).
#'
#' \code{f_keepRegion} is used in \code{\link{data4plot}}.
#' If ReportRegionLevel == RegionSplitLevel, the 2 output region colums will
#' have the same value.
#' NA values and "PEA" are removed silently.
#'
#' @param ReportRegionLevel A string, such as "RegionLevel0".
#' @param ReportRegionValue A string, such as "EMEAR".
#' @param RegionSplitLevel A string, such as "RegionLevel2".
#' @param data A dataframe with columns RegionLevel0, 1, 2, 4.
#' @return a dataframe with columns RegionReport, RegionSplit.
f_keepRegion <- function(data, ReportRegionLevel, ReportRegionValue, RegionSplitLevel) {
regionFacet <- (ReportRegionLevel == RegionSplitLevel)
ReportRegionLevel <- dplyr::sym(ReportRegionLevel)
RegionSplitLevel <- dplyr::sym(RegionSplitLevel)
data %>%
dplyr::mutate(RegionSplit = !!RegionSplitLevel) %>%
dplyr::mutate(temp = !!ReportRegionLevel) %>%
# take into account the case where RRL == RSL (no facet on region)
dplyr::mutate(RegionReport = if (regionFacet) RegionSplit else temp) %>%
dplyr::filter(RegionReport == ReportRegionValue,
RegionSplit != "PEA",
!is.na(RegionSplit)) %>%
dplyr::select_at(dplyr::vars(-dplyr::matches("^RegionLevel|temp")))
}
#' Keep the relevant supplies columns (1 or 2 distinct).
#'
#' \code{f_keepSupplies} is used in \code{\link{data4plot}}.
#' If SuppliesPlotLevel == SuppliesSplitLevel, the 2 output supplies columns
#' are identical.
#' NA values are removed silently.
#'
#' @param data a dataframe with columns SuppliesLevel0, 1, 2, 3, 4.
#' @param SuppliesPlotLevel a string, such as "SuppliesLevel0".
#' @param SuppliesPlotValue a string, such as "LFP aqueous ink".
#' @param SuppliesSplitLevel a string, such as "SuppliesLevel3".
#' @return a dataframe with columns SuppliesPlot, SuppliesSplit.
f_keepSupplies <- function(data, SuppliesPlotLevel, SuppliesPlotValue, SuppliesSplitLevel) {
SuppliesPlotLevel <- dplyr::sym(SuppliesPlotLevel)
SuppliesSplitLevel <- dplyr::sym(SuppliesSplitLevel)
data %>%
dplyr::mutate(SuppliesSplit = !!SuppliesSplitLevel,
SuppliesPlot = !!SuppliesPlotLevel) %>%
dplyr::filter(!is.na(SuppliesSplit),
SuppliesPlot == SuppliesPlotValue) %>%
dplyr::select_at(dplyr::vars(-dplyr::matches("^SuppliesLevel")))
}
#' Rank Region according to revenue (or volume) within the reference year.
#'
#' \code{f_rkRegion} is used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns: fDate, RegionSplit, Rev, Vol.
#' @param yearRef an Integer representing the reference year.
#' @param rankBy a string representing the measure on which the Region are
#' ranked, "Rev" or "Vol".
#' @return a dataframe with an ordered factor for column RegionSplit and
#' column fYear.
f_rkRegion <- function(data, yearRef = 2017, rankBy = "Rev") {
rankBy <- dplyr::sym(rankBy)
data2 <-
data %>%
dplyr::mutate(fYear = lubridate::year(fDate)) %>%
dplyr::group_by(RegionSplit) %>%
dplyr::mutate(Region_rk = sum(ifelse(fYear == yearRef, (!!rankBy), 0), na.rm = T)) %>%
dplyr::ungroup()
# pipeline fail for intance for regionreportvalue=="EFS" and inkreportvalue=="G&J", need to break here
if (nrow(data2) == 0) return("no data f_rkRegion")
data2 %>%
dplyr::mutate(RegionSplit = forcats::fct_reorder(RegionSplit, Region_rk, .desc = TRUE)) %>%
dplyr::select(-Region_rk)
}
#' Rank supplies according to revenue (or volume) within the reference year.
#'
#' \code{f_rkSupplies} is used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns: fYear, SuppliesSplit, SuppliesPlot,
#' Rev, Vol.
#' @param yearRef an Integer representing the reference year.
#' @param rankBy a string representing the measure on which the Region are
#' ranked, "Rev" or "Vol".
#' @return a dataframe with an ordered factor for column SuppliesSplit
#' and fYear.
f_rkSupplies <- function(data, yearRef = 2017, rankBy = "Rev") {
# the case of "no data ..."
if (is.character(data)) return(data)
rankBy <- dplyr::sym(rankBy)
data2 <-
data %>%
dplyr::group_by(SuppliesPlot, SuppliesSplit) %>%
dplyr::mutate(supplies_rk = sum(ifelse(fYear == yearRef, (!!rankBy), 0), na.rm = T)) %>%
dplyr::ungroup()
if (nrow(data2) == 0) return("no data f_rkSupplies")
data2 %>%
dplyr::mutate(rk = dplyr::dense_rank(dplyr::desc(supplies_rk)),
SuppliesSplit = paste0(rk, "-", SuppliesSplit),
SuppliesSplit = forcats::fct_reorder(
SuppliesSplit, supplies_rk, .desc = TRUE)
) %>%
dplyr::select(-supplies_rk)
}
#' Keep the topN supplies items and lump together the Others.
#'
#' \code{f_topSupplies} keeps the topN items on SuppliesSplit column.
#' Used in \code{\link{data4plot}}.
#'
#' @param data A dataframe with columns rk, SuppliesSplit.
#' @param topN An integer representing the number of items to display, the
#' others will be grouped together.
#' @return a dataframe with a column SuppliesSplit, an ordered factor.
f_topSupplies <- function(data, topN = 10) {
# the case of "no data ..."
if (is.character(data)) return(data)
#above, return(data %>% dplyr::select(-rk)), will generate:
#no applicable method for 'select_' applied to an object of class "character"
if (is.null(topN)) return(data)
data %>%
{dplyr::bind_rows(dplyr::filter(., rk <= topN),
dplyr::filter(., rk > topN) %>%
dplyr::mutate(SuppliesSplit = "Others", rk = topN+1)
# NO NEED TO summarise (it's done later)
# %>% group_by(SuppliesSplit, fYear, RegionSplit, rk) %>%
# summarise_if(is.numeric, sum, na.rm = T) %>%
# ungroup
) %>%
dplyr::mutate(SuppliesSplit = forcats::fct_reorder(
SuppliesSplit, rk, .desc = F)) %>%
dplyr::select(-rk)
# ,label = paste0(round(Liters / roundMultiple, ifelse(abs(Liters) < roundThreshold, 1, 0))))
}
}
#' Removes a column to a dataframe.
#'
#' \code{f_keepDateYear} removes a column to a dataframe.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe
#' @param ditch a string, a column name.
#' @return a dataframe without the column ditch.
f_keepDateYear <- function(data, ditch = "fYear") {
# the case of "no data ..."
if (is.character(data)) return(data)
ditch <- dplyr::sym(ditch)
data %>% dplyr::select(-(!!ditch))
}
#' Summarise all numeric columns (sum)
#'
#' \code{f_summariseIFnum} is grouping by the non numeric columns and
#' summarises, with a sum, the numeric columns. If there is a column fYear
#' the column is converted in text during the calculation and return as an
#' integer.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe
#' @return a dataframe
f_summariseIFnum <- function(data) {
# the case of "no data ..."
if (is.character(data)) return(data)
# mutate fYear as character for column selection (group_by_if)
if ("fYear" %in% names(data)) {
data <- data %>% dplyr::mutate(fYear = as.character(fYear))
}
data2 <-
data %>%
# numeric can be Rev, Rev_sm, Liters, Liters_sm
dplyr::group_by_if(~ !is.numeric(.x)) %>%
dplyr::summarise_if(is.numeric, sum, na.rm = T) %>%
dplyr::ungroup()
if ("fYear" %in% names(data)) {
data2 <- data2 %>% dplyr::mutate(fYear = as.integer(fYear))
}
return(data2)
}
#' Prepare data for lollipop plots.
#'
#' \code{f_summarise4lollipop} adds information at level (RegionSplit, fYear),
#' and level (RegionSplit, SuppliesSplit, fYear), while keeping all the rows of
#' the data frame.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with column fYear, lollipop_measure, regionSplit,
#' SuppliesSplit
#' @param lollipop_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a lollipop.
#' @param showFY a string vector, such as "All" or c("2017", "2018").
#' @return a dataframe with columns RegionSplit, SuppliesSplit, fYear,
#' percentage, measure.
f_summarise4lollipop <- function(data, lollipop_measure = "Rev", showFY) {
# lollipop_measure: Null if not lollipop chart, Rev if based on revenue, Liters if based on volume (Liters, sqm)
# the case of "no data ..."
if (is.character(data)) return(data)
if (is.null(lollipop_measure)) return(data)
lollipop_measure <- dplyr::sym(lollipop_measure)
data %>%
dplyr::filter(if (showFY != "All") fYear %in% showFY else TRUE) %>%
dplyr::mutate(measure = !!lollipop_measure) %>%
dplyr::group_by(RegionSplit, fYear) %>%
dplyr::mutate(tot = sum(measure, na.rm = T)) %>%
dplyr::ungroup() %>%
dplyr::group_by(RegionSplit, SuppliesSplit, fYear) %>%
dplyr::mutate(percentage = sum(measure, na.rm = T) / tot) %>%
dplyr::ungroup() %>%
# remove tot, x_sm (and the one not selected among RevVol)
dplyr::select_at(dplyr::vars(-dplyr::matches("^tot$|_sm$")))
}
#' Prepare data for time series plots.
#'
#' \code{f_summarise4ts} filters out old data ans a few old models.
#' Identify the column measure.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns fDate, ts_measure, SuppliesSplit.
#' @param ts_measure a string, such as "Rev_sm" or "Liters_sm". NULL if not a
#' ts plot.
#' @return a dataframe with column fDate, measure, SuppliesSplit.
f_summarise4ts <- function(data, ts_measure = "Rev_sm") {
# the case of "no data ..."
if (is.character(data)) return(data)
if (is.null(ts_measure)) return(data)
# remove the (window-1)th first points of the smoothed values
date2 <- "2014/05/03"
ts_measure <- dplyr::sym(ts_measure)
data2 <-
data %>%
dplyr::filter(# DELETE fDate < lubridate::ymd(date1),
fDate > lubridate::ymd(date2),
!(stringr::str_detect(SuppliesSplit,
"T549_10600|T46X_7000|T47X_9500|T480_7500|T486_5500|T500_10000|T40X_9000"))) %>%
dplyr::mutate(measure = !!ts_measure)
if (nrow(data2) == 0) return("no data f_summarise4ts")
return(data2)
# MUST BE DONE ALREADY IN f_rkSupplies
# group_by(SuppliesPlot, SuppliesSplit, fYear, RegionReport, RegionSplit, fDate) %>%
# summarise(Liters = sum(Liters.sm, na.rm = T),
# LitersFY = sum(Liters, na.rm = T)) %>%
# ungroup %>%
# group_by(SuppliesPlot, SuppliesSplit) %>%
# mutate(ink_rk = sum(Liters, na.rm = T)) %>%
# ungroup
# if (nrow(IS1) == 0) return("no data")
#
# IS1 <- IS1 %>%
# mutate(rk = dense_rank(desc(ink_rk)),
# SuppliesSplit = paste0(rk, "-", SuppliesSplit),
# SuppliesSplit = forcats::fct_reorder(SuppliesSplit, ink_rk, .desc = TRUE)
# # InkLevel = paste0(as.character(as.numeric(InkLevel)), "-", InkLevel)
# ) %>%
# select(-ink_rk, rk)
# ADD THOSE INFORMATION LATER
# find region_RHS, place a label with the inkLevel name at the end of the line of the most RHS region
# also compute the % per region, to be included in the strip.text of the facet
# IS2 <- IS1 %>%
# filter(fYear == yearRef) %>%
# group_by(RegionSplit) %>%
# summarise(LitersFY = sum(LitersFY, na.rm = T)) %>%
# ungroup %>%
# arrange(desc(LitersFY)) %>%
# mutate(perc = 100 * LitersFY/sum(LitersFY))
# region_RHS <- IS2 %>%
# #first, wrapper around [[
# select(RegionSplit) %>% unlist %>% last
#
# IS1 <- IS1 %>%
# group_by(RegionSplit) %>%
# mutate(label = if_else(fDate == max(fDate) & RegionSplit == region_RHS, as.character(as.numeric(SuppliesSplit)), NA_character_))
#
# add in the strip.text of the facet the % represented by the region
# if (length(levels(IS1$RegionSplit)) != nrow(IS2)) print("inkLiLj. Error with number of levels in region_RHS and yearRef")
# if (length(levels(IS1$RegionSplit))>1) levels(IS1$RegionSplit) <- paste0(levels(IS1$RegionSplit), "\n", round(IS2$perc,0), "% in FY", yearRef)
}
#' Prepare data for mosaic plots.
#'
#' \code{f_summarise4mosaic} filters the fiscal years, identifies the measure
#' column.
#' Used in \code{\link{data4plot}}.
#'
#' @param data a dataframe with columns source, RegionSplit, RegionReport,
#' SuppliesSplit, SuppliesPlot, fYear, mosaic_measure.
#' @param mosaic_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a mosaic.
#' @param showFY a string vector, such as "All" or c("2017", "2018")
#' @return a dataframe with columns source, RegionSplit, RegionReport,
#' SuppliesSplit, SuppliesPlot, fYear, measure.
f_summarise4mosaic <- function(data, mosaic_measure, showFY) {
# the case of "no data ..."
if (is.character(data)) return(data)
if (is.null(mosaic_measure)) return(data)
mosaic_measure <- dplyr::sym(mosaic_measure)
data %>%
dplyr::mutate(measure = !!mosaic_measure) %>%
dplyr::filter(if (showFY != "All") fYear %in% showFY else TRUE) %>%
dplyr::select(source, RegionSplit, RegionReport, SuppliesSplit, SuppliesPlot, fYear, measure)
}
#' Prepare data for RVP plots.
#'
#' \code{f_summarise4RVP} adds information at level (RegionSplit, fYear),
#' and level (RegionSplit, SuppliesSplit, fYear), while keeping all the rows of
#' the data frame.
#' Requires to drop fYear before (\code{f_keepDateYear}).
#'
#' @param data a dataframe with columns SuppliesSplit, RegionReport, Supplies,
#' facetY, Rev, Liters, Rev_sm, Liters_sm, Date, fDate.
#' @param facetY a string, such as "RegionSplit", "source". NULL is the plot
#' isn't an RVP.
#' @return a dataframe with columns SuppliesSplit, RegionReport, Supplies,
#' facetY, KEY, VALUE, Date, fDate, label. KEY (Rev_sm, Liters_sm, price) a
#' factor.
f_summarise4RVP <- function(data, facetY) {
if (is.null(facetY)) return(data)
strip_sep <- ifelse(facetY == "size", "\n", "*")
# cat(file = stderr(), "f_summarise4RVP > facetY: ", facetY, ", selected sep: ", strip_sep, "\n")
# remove the first (window-1)th month
date2 <- "2014/05/03"
data %>%
dplyr::filter(fDate > lubridate::ymd(date2),
!(stringr::str_detect(SuppliesSplit,
"T549_10600|T46X_7000|T47X_9500|T480_7500|T486_5500|T500_10000|T40X_9000"))) %>%
dplyr::mutate(facetY = !!dplyr::sym(facetY)) %>%
dplyr::select(-Rev, -Liters) %>%
dplyr::group_by_at(dplyr::vars(dplyr::matches(
"facetY|RegionReport|Supplies|Rev_sm|Liters_sm|Date"))) %>%
dplyr::summarise_if(is.numeric, dplyr::funs(sum(., na.rm = T))) %>%
dplyr::ungroup() %>%
dplyr::mutate(price = ifelse(Liters_sm == 0 | is.na(Liters_sm), 0, Rev_sm / Liters_sm)) %>%
tidyr::gather(key = KEY, value = VALUE, Rev_sm, Liters_sm, price) %>%
dplyr::mutate(KEY = factor(KEY, c("Rev_sm", "Liters_sm", "price"))) %>%
# size = factor(size, c("RA4", "R13", "R24", "R36", "R44", "R64", "SA4", "SA3", "SA2", "SA1"))
#factor created in the order of appearance wit forcarts::as_factor
dplyr::arrange(KEY, facetY) %>%
dplyr::mutate(label = paste0(KEY, strip_sep, facetY),
label = forcats::as_factor(label))
}
# ===< test ===
# dat <- read_csv2("./temp/inkSalesEnriched_RVP.csv", col_types = "ccDccccccccccdddd")
# dat2 <- dat %>%
# f_keepSize(TRUE) %>%
# f_keepSource(FALSE) %>%
# f_keepRegion(ReportRegionLevel = "RegionLevel2", ReportRegionValue = "EU5-NB", RegionSplitLevel = "RegionLevel4") %>%
# f_keepSupplies(SuppliesPlotLevel = "SuppliesLevel1", SuppliesPlotValue = "LFPmedia", SuppliesSplitLevel = "SuppliesLevel2") %>%
# f_rkRegion(yearRef = 2017, rankBy = "Rev") %>%
# f_rkSupplies(yearRef = 2017, rankBy = "Rev") %>%
# f_topSupplies(topN = 10) %>%
# f_keepDateYear(ditch = "fYear") %>%
# f_summariseIFnum %>%
# f_summarise4lollipop(lollipop_measure = NULL, showFY = "All") %>%
# f_summarise4ts(ts_measure = NULL) %>%
# f_summarise4mosaic(mosaic_measure = NULL, showFY = "All") %>%
# f_summarise4RVP(facetY = "size")
# -----
# lollipop_input <- data4plot_Lollipop(dat, RRL = "RegionLevel2", RRV = "EU5-NB", RSL = "RegionLevel2",
# SPL = "SuppliesLevel1", SPV = "aqueous", SSL = "SuppliesLevel2",
# yearRef = 2017, rankBy = "Rev", topN = 10,
# lollipop_measure = "Rev")
# p <- plot_lollipop(lollipop_input$data, lollipop_input$labs)
# -----
# ts_input <- data4plot_ts(dat, RRL="RegionLevel2", RRV="EU5-NB", RSL="RegionLevel2",
# SPL="SuppliesLevel1", SPV="aqueous", SSL="SuppliesLevel2",
# yearRef=2017, rankBy="Rev", topN=10, ts_measure="Rev")
# plot_ts(ts_input$data, ts_input$labs)
# -----
# mosaic_input <- data4plot_Mosaic(dat, RRL="RegionLevel2", RRV="EU5-NB", RSL="RegionLevel2",
# SPL="SuppliesLevel1", SPV="aqueous", SSL="SuppliesLevel2",
# yearRef=2017, showFY=2017, rankBy="Rev", topN=5, mosaic_measure="Rev")
# plot_mosaic(mosaic_input$data, mosaic_input$labs)
# plot_mosaic(dat2, c("labs1", "labs2", "labs3")) %>% ggplotly
# === end test >===
#' Prepare data for plot lollipop, ts, mosaic, RVP.
#'
#' \code{data4plot} regroups several functions, in order:
#' \code{\link{f_keepSize}}, \code{\link{f_keepSource}},
#' \code{\link{f_keepRegion}}, \code{\link{f_keepSupplies}},
#' \code{\link{f_rkRegion}}, \code{\link{f_rkSupplies}},
#' \code{\link{f_topSupplies}}, \code{\link{f_keepDateYear}},
#' \code{\link{f_summariseIFnum}}, \code{\link{f_summarise4lollipop}},
#' \code{\link{f_summarise4ts}}, \code{\link{f_summarise4mosaic}},
#' \code{\link{f_summarise4RVP}}.
#' \code{data4plot} is used in:
#' \code{\link{data4plot_RVP}}
#' \code{\link{data4plot_Lollipop}}
#' \code{\link{data4plot_Mosaic}}
#' \code{\link{data4plot_ts}}
#'
#' @param data a dataframe.
#' @param keepSize a boolean.
#' @param keepSource a boolean.
#' @param RRL a string, reportRegionLevel.
#' @param RRV a string, ReportRegionValue.
#' @param RSL a string, RegionSplitLevel.
#' @param SPL a string, SuppliesPlotLevel.
#' @param SPV a string, SuppliesPlotValue.
#' @param SSL a string, SppliesSplitLevel.
#' @param yearRef an Integer representing the reference year.
#' @param rankBy a string representing the measure on which the Region are
#' ranked, "Rev" or "Vol".
#' @param topN An integer representing the number of items to display, the
#' others will be grouped together.
#' @param ditch a string, a column name.
#' @param showFY a string vector, such as "All" or c("2017", "2018").
#' @param lollipop_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a lollipop.
#' @param ts_measure a string, such as "Rev_sm" or "Liters_sm". NULL if not a
#' ts plot.
#' @param mosaic_measure a string, such as "Rev" or "Liters". NULL if the
#' plot isn't a mosaic.
#' @param RVP_facetY a string, such as "RegionSplit", "source". NULL is the
#' plot isn't an RVP.
#' @return a dataframe expected by one of the plot functions.
data4plot <- function(data, keepSize, keepSource,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch, showFY,
lollipop_measure, ts_measure, mosaic_measure, RVP_facetY) {
# DEBUG
# cat(file = stderr(), "data4plot > RP_facetY: ", RVP_facetY, "\n")
data2 <- data %>%
f_keepSize(keepSize) %>%
f_keepSource(keepSource) %>%
f_keepRegion(ReportRegionLevel = RRL,
ReportRegionValue = RRV,
RegionSplitLevel = RSL) %>%
f_keepSupplies(SuppliesPlotLevel = SPL,
SuppliesPlotValue = SPV,
SuppliesSplitLevel = SSL) %>%
f_rkRegion(yearRef, rankBy) %>%
f_rkSupplies(yearRef, rankBy) %>%
f_topSupplies(topN) %>%
f_keepDateYear(ditch) %>%
f_summariseIFnum %>%
f_summarise4lollipop(lollipop_measure, showFY) %>%
f_summarise4ts(ts_measure) %>%
f_summarise4mosaic(mosaic_measure, showFY) %>%
f_summarise4RVP(RVP_facetY)
return(data2)
}
#' Returns data for plot_RVP.
#'
#' \code{data4plot_RVP} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @param free_y a string, such as "free_y".
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_RVP <- function(data, keepSize, keepSource, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN,
RVP_facetY,
SP_filter_part, free_y) {
data2 <- data4plot(data, keepSize, keepSource,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fYear", showFY = "All",
lollipop_measure = NULL, ts_measure = NULL,
mosaic_measure = NULL, RVP_facetY)
labs <- list(RVP_facetY = RVP_facetY, SPV = SPV, SP_filter_part = SP_filter_part, free_y = free_y)
return(list(data = data2, labs = labs))
}
#' Returns the order or magnitude of a double (1, 1 000, 100 000, ...).
#'
#' \code{divid} is used \code{\link{data4plot_Lollipop}}.
#'
#' @param y a double.
#' @return an integer representing the order of magnitude.
divid <- function(y) {
signif(y,3) %>% log10 %>% `%/%`(3) %>% floor %>% (function(x) {10^(3*x)})
}
#' Returns data for plot_lollipop.
#'
#' \code{data4plot_Lollipop} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_Lollipop <- function(data, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, showFY,
lollipop_measure,
SP_filter_part) {
data2 <- data4plot(data, keepSize = F, keepSource = F,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fDate", showFY,
lollipop_measure, ts_measure = NULL,
mosaic_measure = NULL, RVP_facetY = NULL)
div <- data2 %>%
dplyr::select(measure) %>% unlist %>% max(na.rm = T) %>% divid
vol_unit <- data_SL2_regex[data_SL2_regex$SuppliesLevel2_type == SP_filter_part, "unit"][[1]]
measure_txt <- ifelse(stringr::str_detect(lollipop_measure, "Rev"), "Revenue", "Volume")
vol_unit <- ifelse(measure_txt == "Revenue", "Euros NetNet", vol_unit)
labs <- list(title = paste0(measure_txt,
" (x ", format(div, scientific = FALSE, big.mark = " "), " ", vol_unit,
") per FY. Rank according to FY", yearRef),
subtitle = paste0("sell-in data for the top ", topN, " models in product group ", SPV, "."),
x_lab = "% of total",
y_lab = "unit",
not_use1 = "Ink / Printer model",
caption = paste0(RRL, RRV, RSL,"_", SPL, SPV, SSL, "_", yearRef,
lollipop_measure, collapse = "_"),
div = div)
# labs <- c(paste0("Volume (x", div," sqm or Liters) per FY. Rank according to FY", yearRef),
# paste0("sell-in data for the top ", topN, " models in product group ", SPV, "."),
# "% of total ",
# "unit",
# "Ink / Printer model",
# paste0(RRL, RRV, RSL,"_", SPL, SPV, SSL, "_", yearRef,
# lollipop_measure, collapse = "_"),
# div)
return(list(data = data2, labs = labs))
}
#' Returns data for plot_mosaic.
#'
#' \code{data4plot_Mosaic} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_Mosaic <- function(data, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, showFY,
mosaic_measure,
SP_filter_part) {
data2 <- data4plot(data, keepSize = F, keepSource = T,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fDate", showFY,
lollipop_measure = NULL, ts_measure = NULL,
mosaic_measure, RVP_facetY = NULL)
vol_unit <- data_SL2_regex[data_SL2_regex$SuppliesLevel2_type == SP_filter_part, "unit"][[1]]
measure_txt <- ifelse(stringr::str_detect(mosaic_measure, "Rev"), "Revenue", "Volume")
vol_unit <- ifelse(measure_txt == "Revenue", "Euros NetNet", vol_unit)
labs <- c(paste0("Split of ", measure_txt, " (", vol_unit, "), per categories (within ", SPV, ") and channel ", RRV),
paste0("Top", topN, " categories | QTs are QTs approved by EMD"),
"plotNameMosaic")
#test
write.csv2(data2, "test_mosaic_jan18.csv")
return(list(data = data2, labs = labs))
}
#' Returns data for plot_ts.
#'
#' \code{data4plot_ts} uses \code{\link{data4plot}}.
#'
#' @param SP_filter_part a string helping to selct a regular expression in the
#' dataframe data_SL2_regex.
#' @param free_y a string, such as "free_y".
#' @return a list containing a dataframe, data, and a string vector, labs.
data4plot_ts <- function(data, RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN,
ts_measure,
SP_filter_part, free_y) {
data2 <- data4plot(data, keepSize = F, keepSource = F,
RRL, RRV, RSL,
SPL, SPV, SSL,
yearRef, rankBy, topN, ditch = "fYear", showFY = "All",
lollipop_measure = NULL, ts_measure,
mosaic_measure = NULL, RVP_facetY = NULL)
labs <- list(SPV = SPV, SP_filter_part = SP_filter_part, ts_measure = ts_measure, free_y = free_y)
return(list(data = data2, labs = labs))
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.