library(presenter)
library(dplyr)
Output is automatically formatted depending on the mode of the columns
When using make excel, the first input is a dataframe, but do not input an expression or use a pipe. A variable name is necessary for proper naming of the excel output, which is sent to the working directory and named after the data frame.
The recommended ouput format is to put ID style columns first and and value columns after. Last_id_col takes the integer index of the last id column.
If value columns are organized by header words they can be included in a character string. This will color code the column headers. Split and merge cells are not used to maintain ease of read / write compatibility of the workbook with R.
make_excel(df = iris,
header_word = c("Sepal", "Petal"),
last_id_col = NULL)
make_simple_excel
allows the user to export to excel but
with minimal formatting. THe data frame can be piped in, and unless
specified otherwise in `output_file, the resulting excel file will be
named after the data frame.
%>%
iris make_simple_excel()
make_simple_excel
also accepts a list of dfs, and
exports one per sheet. This is useful when you automatically generated a
list of tables, or saved a bunch of table summaries in a list.
%>%
iris group_by(Species) %>%
summarize(across(where(is.numeric), sum)) -> iris_summary
list(iris, iris_summary) -> iris_list
%>%
iris_list make_simple_excel()
Use the automated formatting paradigm and customize the id cols and header words for each sheet.
make_excel_wb(wb = NULL,
object = iris,
last_id_col = NULL,
header_word = c("Sepal", "Petal")) %>%
make_excel_wb(object = anscombe,
last_id_col = NULL,
header_word = NULL) %>%
finish_excel_wb(wb_name = "data_workbook")