diff --git a/DESCRIPTION b/DESCRIPTION
index 7a5ff1a3..7de12b7f 100755
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
Package: AMR
-Version: 0.2.0.9011
-Date: 2018-07-15
+Version: 0.2.0.9012
+Date: 2018-07-16
Title: Antimicrobial Resistance Analysis
Authors@R: c(
person(
diff --git a/NAMESPACE b/NAMESPACE
index 5787295f..c143ac82 100755
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -8,6 +8,7 @@ S3method(as.vector,frequency_tbl)
S3method(as_tibble,frequency_tbl)
S3method(barplot,mic)
S3method(barplot,rsi)
+S3method(format,frequency_tbl)
S3method(hist,frequency_tbl)
S3method(kurtosis,data.frame)
S3method(kurtosis,default)
@@ -74,6 +75,7 @@ exportMethods(as.vector.frequency_tbl)
exportMethods(as_tibble.frequency_tbl)
exportMethods(barplot.mic)
exportMethods(barplot.rsi)
+exportMethods(format.frequency_tbl)
exportMethods(hist.frequency_tbl)
exportMethods(kurtosis)
exportMethods(kurtosis.data.frame)
@@ -97,12 +99,15 @@ importFrom(clipr,read_clip_tbl)
importFrom(clipr,write_clip)
importFrom(curl,nslookup)
importFrom(dplyr,"%>%")
+importFrom(dplyr,all_vars)
+importFrom(dplyr,any_vars)
importFrom(dplyr,arrange)
importFrom(dplyr,arrange_at)
importFrom(dplyr,as_tibble)
importFrom(dplyr,between)
importFrom(dplyr,desc)
importFrom(dplyr,filter)
+importFrom(dplyr,filter_at)
importFrom(dplyr,group_by)
importFrom(dplyr,group_by_at)
importFrom(dplyr,if_else)
@@ -118,6 +123,7 @@ importFrom(dplyr,slice)
importFrom(dplyr,summarise)
importFrom(dplyr,tibble)
importFrom(dplyr,top_n)
+importFrom(dplyr,vars)
importFrom(grDevices,boxplot.stats)
importFrom(graphics,axis)
importFrom(graphics,barplot)
diff --git a/NEWS.md b/NEWS.md
index 27d5fe6f..7deef81d 100755
--- a/NEWS.md
+++ b/NEWS.md
@@ -14,7 +14,7 @@ ratio(c(772, 1611, 737), ratio = "1:2:1")
* A vignette to explain its usage
* Support for `table` to use as input: `freq(table(x, y))`
* Support for existing functions `hist` and `plot` to use a frequency table as input: `hist(freq(df$age))`
- * Support for `as.vector`, `as.data.frame` and `as_tibble`
+ * Support for `as.vector`, `as.data.frame`, `as_tibble` and `format`
* Support for quasiquotation: `freq(mydata, mycolumn)` is the same as `mydata %>% freq(mycolumn)`
* Function `top_freq` function to return the top/below *n* items as vector
* Header of frequency tables now also show Mean Absolute Deviaton (MAD) and Interquartile Range (IQR)
diff --git a/R/clipboard.R b/R/clipboard.R
index bbe95558..d199daf7 100644
--- a/R/clipboard.R
+++ b/R/clipboard.R
@@ -7,6 +7,7 @@
#' This also supports automatic column type transformation, with AMR classes \code{\link{as.rsi}} and \code{\link{as.mic}}.
#' @rdname clipboard
#' @name clipboard
+#' @inheritParams base::data.frame
#' @inheritParams utils::read.table
#' @inheritParams utils::write.table
#' @inheritParams readr::locale
@@ -20,6 +21,10 @@
#' @importFrom utils read.delim write.table object.size
#' @importFrom readr parse_guess locale
#' @details
+#' The parameter \code{stringsAsFactors} defaults to \code{FALSE}, as opposed to most base \R methods.
+#'
+#' The parameters \code{date_format} and \code{time_format} also support generic date and time formats like \code{"dd-mm-yyyy"} like Excel.
+#'
#' \if{html}{
#' \strong{Example for copying from Excel:}
#' \out{}\figure{clipboard_copy.png}\out{
}
@@ -36,10 +41,27 @@
#' \out{}\figure{clipboard_rsi.png}\out{
}
#' }
#' @export
+#' @examples
+#' \dontrun{
+#'
+#' df1 <- data.frame(a = letters[1:12],
+#' b = runif(n = 12, min = 1000, max = 2000),
+#' stringsAsFactors = FALSE)
+#' clipboard_export(df1)
+#' df2 <- clipboard_import()
+#' identical(df1, df2)
+#'
+#' # send frequency table to clipboard (e.g. for pasting in Excel)
+#' septic_patients %>%
+#' freq(age) %>%
+#' format() %>% # this will format the percentages
+#' clipboard_export()
+#' }
clipboard_import <- function(sep = '\t',
header = TRUE,
dec = ".",
na = c("", "NA", "NULL"),
+ stringsAsFactors = FALSE,
startrow = 1,
as_vector = TRUE,
guess_col_types = TRUE,
@@ -58,7 +80,7 @@ clipboard_import <- function(sep = '\t',
dec = dec,
na.strings = na,
encoding = 'UTF-8',
- stringsAsFactors = FALSE)
+ stringsAsFactors = stringsAsFactors)
# use tibble, so column types will be translated correctly
import_tbl <- as_tibble(import_tbl)
@@ -91,6 +113,9 @@ clipboard_import <- function(sep = '\t',
import_tbl <- import_tbl %>% pull(1)
}
+ # and transform back to data.frame
+ import_tbl <- as.data.frame(import_tbl, stringsAsFactors = stringsAsFactors)
+
if (info == TRUE) {
cat("Successfully imported from clipboard:", NROW(import_tbl), "obs. of", NCOL(import_tbl), "variables.\n")
}
diff --git a/R/freq.R b/R/freq.R
index f73640ef..9d88c560 100755
--- a/R/freq.R
+++ b/R/freq.R
@@ -100,10 +100,13 @@
#' # show only the top 5
#' years %>% print(nmax = 5)
#'
+#' # save to an object with formatted percentages
+#' years <- format(years)
+#'
#' # print a histogram of numeric values
#' septic_patients %>%
#' freq(age) %>%
-#' hist() # prettier: ggplot(septic_patients, aes(age)) + geom_histogram()
+#' hist() # prettier: ggplot(septic_patients, aes(age)) + geom_histogram()
#'
#' # or print all points to a regular plot
#' septic_patients %>%
@@ -124,13 +127,20 @@
#' freq(age) %>%
#' as.vector() %>%
#' sort(),
-#' sort(septic_patients$age)
-#' ) # TRUE
+#' sort(septic_patients$age)) # TRUE
#'
-#' # also supports table:
+#' # it also supports `table` objects:
#' table(septic_patients$sex,
#' septic_patients$age) %>%
-#' freq()
+#' freq(sep = " **sep** ")
+#'
+#' \dontrun{
+#' # send frequency table to clipboard (e.g. for pasting in Excel)
+#' septic_patients %>%
+#' freq(age) %>%
+#' format() %>% # this will format the percentages
+#' clipboard_export()
+#' }
frequency_tbl <- function(x,
...,
sort.count = TRUE,
@@ -603,3 +613,20 @@ plot.frequency_tbl <- function(x, y, ...) {
as.vector.frequency_tbl <- function(x, mode = "any") {
as.vector(rep(x$item, x$count), mode = mode)
}
+
+#' @noRd
+#' @exportMethod format.frequency_tbl
+#' @export
+format.frequency_tbl <- function(x, digits = 1, ...) {
+ opt <- attr(x, 'opt')
+ if (opt$nmax.set == TRUE) {
+ nmax <- opt$nmax
+ } else {
+ nmax <- getOption("max.print.freq", default = 15)
+ }
+
+ x <- x[1:nmax,]
+ x$percent <- percent(x$percent, round = digits, force_zero = TRUE)
+ x$cum_percent <- percent(x$cum_percent, round = digits, force_zero = TRUE)
+ base::format.data.frame(x, ...)
+}
diff --git a/man/clipboard.Rd b/man/clipboard.Rd
index 945853cf..e730ab56 100644
--- a/man/clipboard.Rd
+++ b/man/clipboard.Rd
@@ -7,9 +7,10 @@
\title{Import/export from clipboard}
\usage{
clipboard_import(sep = "\\t", header = TRUE, dec = ".", na = c("", "NA",
- "NULL"), startrow = 1, as_vector = TRUE, guess_col_types = TRUE,
- date_names = "en", date_format = "\%Y-\%m-\%d", time_format = "\%H:\%M",
- tz = Sys.timezone(), encoding = "UTF-8", info = TRUE)
+ "NULL"), stringsAsFactors = FALSE, startrow = 1, as_vector = TRUE,
+ guess_col_types = TRUE, date_names = "en", date_format = "\%Y-\%m-\%d",
+ time_format = "\%H:\%M", tz = Sys.timezone(), encoding = "UTF-8",
+ info = TRUE)
clipboard_export(x, sep = "\\t", dec = ".", na = "", header = TRUE,
info = TRUE)
@@ -30,6 +31,11 @@ clipboard_export(x, sep = "\\t", dec = ".", na = "", header = TRUE,
\item{na}{the string to use for missing values in the data.}
+\item{stringsAsFactors}{logical: should character vectors be converted
+ to factors? The \sQuote{factory-fresh} default is \code{TRUE}, but
+ this can be changed by setting \code{\link{options}(stringsAsFactors
+ = FALSE)}.}
+
\item{startrow}{\emph{n}th row to start importing from. When \code{header = TRUE}, the import will start on row \code{startrow} \emph{below} the header.}
\item{as_vector}{a logical value indicating whether data consisting of only one column should be imported as vector using \code{\link[dplyr]{pull}}. This will strip off the header.}
@@ -80,7 +86,11 @@ The data will be read and written as tab-separated by default, which makes it po
This also supports automatic column type transformation, with AMR classes \code{\link{as.rsi}} and \code{\link{as.mic}}.
}
\details{
-\if{html}{
+The parameter \code{stringsAsFactors} defaults to \code{FALSE}, as opposed to most base \R methods.
+
+The parameters \code{date_format} and \code{time_format} also support generic date and time formats like \code{"dd-mm-yyyy"} like Excel.
+
+ \if{html}{
\strong{Example for copying from Excel:}
\out{}\figure{clipboard_copy.png}\out{
}
\cr
@@ -96,6 +106,23 @@ This also supports automatic column type transformation, with AMR classes \code{
\out{}\figure{clipboard_rsi.png}\out{
}
}
}
+\examples{
+\dontrun{
+
+df1 <- data.frame(a = letters[1:12],
+ b = runif(n = 12, min = 1000, max = 2000),
+ stringsAsFactors = FALSE)
+clipboard_export(df1)
+df2 <- clipboard_import()
+identical(df1, df2)
+
+# send frequency table to clipboard (e.g. for pasting in Excel)
+septic_patients \%>\%
+ freq(age) \%>\%
+ format() \%>\% # this will format the percentages
+ clipboard_export()
+}
+}
\keyword{clipboard}
\keyword{clipboard_export}
\keyword{clipboard_import}
diff --git a/man/freq.Rd b/man/freq.Rd
index f4745802..7bbc2099 100755
--- a/man/freq.Rd
+++ b/man/freq.Rd
@@ -111,10 +111,13 @@ years <- septic_patients \%>\%
# show only the top 5
years \%>\% print(nmax = 5)
+# save to an object with formatted percentages
+years <- format(years)
+
# print a histogram of numeric values
septic_patients \%>\%
freq(age) \%>\%
- hist() # prettier: ggplot(septic_patients, aes(age)) + geom_histogram()
+ hist() # prettier: ggplot(septic_patients, aes(age)) + geom_histogram()
# or print all points to a regular plot
septic_patients \%>\%
@@ -135,13 +138,20 @@ identical(septic_patients \%>\%
freq(age) \%>\%
as.vector() \%>\%
sort(),
- sort(septic_patients$age)
-) # TRUE
+ sort(septic_patients$age)) # TRUE
-# also supports table:
+# it also supports `table` objects:
table(septic_patients$sex,
septic_patients$age) \%>\%
- freq()
+ freq(sep = " **sep** ")
+
+\dontrun{
+# send frequency table to clipboard (e.g. for pasting in Excel)
+septic_patients \%>\%
+ freq(age) \%>\%
+ format() \%>\% # this will format the percentages
+ clipboard_export()
+}
}
\keyword{freq}
\keyword{frequency}