Upgrade from openxlsx

Basic read and write functions

Welcome to the openxlsx2 update vignette. In this vignette we will take some common code examples from openxlsx and show you how similar results can be replicated in openxlsx2. Thank you for taking a look, and let’s get started. While previous openxlsx functions used the . in function calls, as well as camelCase, we have tried to switch to snake_case (this is still a work in progress, there may still be function arguments that use camelCase).

Read xlsx or xlsm files

The basic read function changed from read.xlsx to read_xlsx. Using a default xlsx file included in the package:

file <- system.file("extdata", "openxlsx2_example.xlsx", package = "openxlsx2")

The old syntax looked like this:

# read in openxlsx
openxlsx::read.xlsx(xlsxFile = file)

This has changed to this:

# read in openxlsx2
openxlsx2::read_xlsx(file = file)
##     Var1 Var2 <NA>  Var3  Var4       Var5         Var6    Var7     Var8
## 3   TRUE    1   NA     1     a 2023-05-29 3209324 This #DIV/0! 01:27:15
## 4   TRUE   NA   NA #NUM!     b 2023-05-23         <NA>       0 14:02:57
## 5   TRUE    2   NA  1.34     c 2023-02-01         <NA> #VALUE! 23:01:02
## 6  FALSE    2   NA  <NA> #NUM!       <NA>         <NA>       2 17:24:53
## 7  FALSE    3   NA  1.56     e       <NA>         <NA>    <NA>     <NA>
## 8  FALSE    1   NA   1.7     f 2023-03-02         <NA>     2.7 08:45:58
## 9     NA   NA   NA  <NA>  <NA>       <NA>         <NA>    <NA>     <NA>
## 10 FALSE    2   NA    23     h 2023-12-24         <NA>      25     <NA>
## 11 FALSE    3   NA  67.3     i 2023-12-25         <NA>       3     <NA>
## 12    NA    1   NA   123  <NA> 2023-07-31         <NA>     122     <NA>

As you can see, we return the spreadsheet return codes (e.g., #NUM) in openxlsx2. Another thing to see above, we return the cell row as rowname for the data frame returned. openxlsx2 should return a data frame of the selected size, even if it empty. If you preferred openxlsx::readWorkbook() this has become wb_read(). All of these are wrappers for the newly introduced function wb_to_df() which provides the most options. read_xlsx() and wb_read() were created for backward comparability.

Write xlsx files

Basic writing in openxlsx2 behaves identical to openxlsx. Though be aware that overwrite is an optional parameter in openxlsx2 and just like in other functions like base::write.csv() if you write onto an existing file name, this file will be replaced.

Setting the output to some temporary xlsx file

output <- temp_xlsx()

The previous write function looks like this:

# write in openxlsx
openxlsx::write.xlsx(iris, file = output, colNames = TRUE)

The new function looks quite similar:

# write in openxlsx2
openxlsx2::write_xlsx(iris, file = output, col_names = TRUE)

Basic workbook functions

Workbook functions have been renamed to begin with wb_ there are plenty of these in the package, therefore looking at the man pages seems to be the fastest way. Yet, it all begins with loading the workbook.

Loading a workbook

A major feature in openxlsx are workbooks. Obviously they remain a central piece in openxlsx2. Previous you would load them with:

wb <- openxlsx::loadWorkbook(file = file)

In openxlsx2 loading was changed to:

wb <- wb_load(file = file)

There are plenty of functions to interact with workbooks and we will not describe every single one here. A detailed list can be found over at our references

Styles

One of the biggest user facing change was the removal of the stylesObject. In the following section we use code from openxlsx::addStyle()

# openxlsx
## Create a new workbook
wb <- createWorkbook(creator = "My name here")
addWorksheet(wb, "Expenditure", gridLines = FALSE)
writeData(wb, sheet = 1, USPersonalExpenditure, rowNames = TRUE)

## style for body
bodyStyle <- createStyle(border = "TopBottom", borderColor = "#4F81BD")
addStyle(wb, sheet = 1, bodyStyle, rows = 2:6, cols = 1:6, gridExpand = TRUE)

## set column width for row names column
setColWidths(wb, 1, cols = 1, widths = 21)

In openxlsx2 the same code looks something like this:

# openxlsx2 chained
border_color <- wb_color(hex = "4F81BD")
wb <- wb_workbook(creator = "My name here")$
  add_worksheet("Expenditure", grid_lines = FALSE)$
  add_data(x = USPersonalExpenditure, row_names = TRUE)$
  add_border( # add the outer and inner border
    dims = "A1:F6",
    top_border = "thin", top_color = border_color,
    bottom_border = "thin", bottom_color = border_color,
    inner_hgrid = "thin", inner_hcolor = border_color,
    left_border = "", right_border = ""
  )$
  set_col_widths( # set column width
    cols = 1:6,
    widths = c(20, rep(10, 5))
  )$ # remove the value in A1
  add_data(dims = "A1", x = "")

The code above uses chaining. If you prefer piping, we provide the chained functions with the prefix wb_ so wb_add_worksheet(), wb_add_data(), wb_add_border() and wb_set_col_widths() would be the functions to use with pipes %>% or |>.

With pipes the code from above becomes

# openxlsx2 with pipes
border_color <- wb_color(hex = "4F81BD")
wb <- wb_workbook(creator = "My name here") %>%
  wb_add_worksheet(sheet = "Expenditure", grid_lines = FALSE) %>%
  wb_add_data(x = USPersonalExpenditure, row_names = TRUE) %>%
  wb_add_border( # add the outer and inner border
    dims = "A1:F6",
    top_border = "thin", top_color = border_color,
    bottom_border = "thin", bottom_color = border_color,
    inner_hgrid = "thin", inner_hcolor = border_color,
    left_border = "", right_border = ""
  ) %>%
  wb_set_col_widths( # set column width
    cols = 1:6,
    widths = c(20, rep(10, 5))
  ) %>% # remove the value in A1
  wb_add_data(dims = "A1", x = "")

Be aware that chains modify an object in place and pipes do not.

# openxlsx2
wbp <- wb_workbook() %>% wb_add_worksheet()
wbc <- wb_workbook()$add_worksheet()

# need to assign wbp
wbp <- wbp %>% wb_add_data(x = iris)
wbc$add_data(x = iris)

You can re-use styles with wb_get_cell_style() and wb_set_cell_style(). Abandoning stylesObject in openxlsx2 has the huge benefit that we can import and export a spreadsheet without changing any cell style. It is still possible to modify a cell style with wb_add_border(), wb_add_fill(), wb_add_font() and wb_add_numfmt().

Additional examples regarding styles can be found in the styles vignette.

Conditional formatting

See vignette("conditional-formatting") for extended examples on formatting.

Here is a minimal example:

# openxlsx2 with chains
wb <- wb_workbook()$
  add_worksheet("a")$
  add_data(x = 1:4, col_names = FALSE)$
  add_conditional_formatting(dims = "A1:A4", rule = ">2")

# openxlsx2 with pipes
wb <- wb_workbook() %>%
  wb_add_worksheet("a") %>%
  wb_add_data(x = 1:4, col_names = FALSE) %>%
  wb_add_conditional_formatting(dims = "A1:A4", rule = ">2")

Data validation

Similarly, data validation has been updated and improved. This openxlsx code for data validation

# openxlsx
wb <- createWorkbook()
addWorksheet(wb, "Sheet 1")
writeDataTable(wb, 1, x = iris[1:30, ])
dataValidation(wb, 1,
  col = 1:3, rows = 2:31, type = "whole",
  operator = "between", value = c(1, 9)
)

looks in openxlsx2 something like this:

# openxlsx2 with chains
wb <- wb_workbook()$
  add_worksheet("Sheet 1")$
  add_data_table(1, x = iris[1:30, ])$
  add_data_validation(1,
    dims = wb_dims(rows = 2:31, cols = 1:3),
    # alternatively, dims can also be "A2:C31" if you know the span in your Excel workbook.
    type = "whole",
    operator = "between",
    value = c(1, 9)
  )

# openxlsx2 with pipes
wb <- wb_workbook() %>%
  wb_add_worksheet("Sheet 1") %>%
  wb_add_data_table(1, x = iris[1:30, ]) %>%
  wb_add_data_validation(
    sheet = 1,
    dims = "A2:C31", # alternatively, dims = wb_dims(rows = 2:31, cols = 1:3)
    type = "whole",
    operator = "between",
    value = c(1, 9)
  )

Saving

Saving has been switched from saveWorkbook() to wb_save() and opening a workbook has been switched from openXL() to wb_open().

Why openxlsx2?

Originally, openxlsx2 was started as a private branch of openxlsx to include the pugixml library and provide a fully functional XML parser for openxlsx. At that time, it became clear that the home-written openxlsx XML parser was limited in its ability to reliably parse XML files, leading to some problems with broken and unreadable xlsx files. Once the inclusion of pugixml was addressed, a new internal structure was created, and this structure required changes to most of the old openxlsx functions. This was accompanied by the change from methods to R6 and the possibility of chaining and piping functions.

Working with the styles object of openxlsx it became clear that while it is a great idea, it does not work well enough for our needs and that files loaded and modified by openxlsx never look the same. There are always things that look a little off because the style objects do not work perfectly. Likewise, there are a lot of edge cases in openxlsx that assume a file structure in xlsx objects that is a simplified approximation of what is actually going on. For example, openxlsx assumes that each sheet is accompanied by a drawing. While this works in many cases, it does not match the definition of the format in the openxml standard. There may be worksheets with multiple drawings, and there should be no drawing folder if no drawings are included. Unfortunately, many of these things are deeply embedded in the openxlsx code, and the more development that took place in openxlsx2, the more things differed between the fork from its origin. At some point the fork was called an independent project and the previously privately developed branch was made public.

You could say that this went hand in hand with the modification of the actual project goal. Before, it was about creating a similar looking xlsx file and being able to partially edit it. Now it was about writing an identical xlsx file and just being able to change everything.

Since then most of the internal functions of openxlsx have been cleaned up, fixed and mostly rewritten. The package has developed new ways to handle styles with the styles manager, it provides a full range of style options that would be hard or impossible to include in openxlsx. We have included support for native graphs with mschart and feature the creation of pivot tables. We support more conditional formatting options, we have improved the support for data validation, we have sparklines and form control objects. In addition many of the quirks of the old package have been ironed out. We have switched to a consistent and stable API build on dims and we provide multiple vignettes to document our code and plenty of functions to interact with the openxml format. We provide basic xlsb support and with msoc we have created a package encrypt and decrypt openxml files.

Invitation to contribute

We have put a lot of work into openxls2 to make it useful for our needs, improving what we found useful about openxlsx and removing what we didn’t need. We do not claim to be omniscient about all the things you can do with spreadsheet software, nor do we claim to be omniscient about all the things you can do in openxlsx2. The package is still under active development, though we have reached a semi stable API that will not change until the next major release.

We are quite fond of our little package and invite others to try it out and comment on what they like and of course what they think we are missing or if something doesn’t work. openxlsx2 is a complex piece of software that certainly does not work bug-free, even if we did our best. If you want to contribute to the development of openxlsx2, please be our guest on our Github. Join or open a discussion, post or fix issues or write us a mail.