Currently only html output is possible.

To add LaTeX output we need to "translate" css to latex instructions.

Suggested implementation

Assume we have:

• Data matrix (of characters, only with the columns to be rendered)
• List of formats. The item names are CSS fields, the items are matrices with corresponding CSS values .

In the HTML case we are:

• Merging the CSS formats in a long format string matrix (here )
• Passing the data and the format to htmlTable

For the LaTeX syntax this approach is not good, as formatting wraps values (\textbf{value} for bold).

• Provide the formatting as a sort of matrix of lambda(value) functions that return the latex code. Kind of an array of function(value) sprintf("\\textbf{%s}", value).
• Replace x here with the call of the proper lambda function to the right value (with a mapply call to iterate along all lambdas and values). myfun <- function(fun, y) return fun(y) and then `mapply(FUN=myfun, matrix_of_lambdas, values). The idea is to obtain a matrix of characters with the LaTeX code for each formatted element.

Hi,

If I am exporting data frames to excel, can I show the currency and percentage symbols in the excel spreadsheet on certain columns ($, %)?

Cheers!

When attempting to use this line of code,
condformat::condformat2excel(cf, "hello.xlsx",,T)

I get the following traceback.

6: stop(gettextf("'%s' is not an exported object from 'namespace:%s'", name, getNamespaceName(ns)), call. = FALSE, domain = NA)
5: getExportedValue(pkg, name)
4: xlsx::CellBlock.default
3: condformat2excelsheet(x, sheet)
2: condformat::condformat2excel(cf, "hello.xlsx", , T) at transcriptome_pharmer.r#363
1: heatMapper(geneDF)

This package is awesome, but I've noticed something that may be a bug.

When knitting to PDF, ## NULL appears before the table output. Apparently, this does not happen when knitting to HTML, per comments in this StackOverflow thread I created.

Could you please take a look?

Note: Please see code in StackOverflow question, as the RMD chunk delimiters aren't working with GitHub's code delimiters.

Hi,

condformat works as intended without a loop, but strangely enough breaks when wrapped in one (I'd like to use it for large tables). I've put together a minimal example at:

https://tex.stackexchange.com/questions/516747/for-loop-breaks-condformat

Cheers.

Check the following case and output:

# Custom discrete color values can be specified with a function. The function takes
# the whole column and returns a vector with the colours.
color_pick <- function(col) {
  ifelse(col < 4.7,
         "yellow",
         ifelse(col < 4.9,
                "orange",
                "purple"))
}
condformat(head(iris)) + rule_fill_discrete_("Sepal.Length", ~ color_pick(Sepal.Length))

In other words, it appears the thresholds are being respected, but the colors being passed are not.

Apart from LaTeX and HTML, it would be cool to have a condformat2grob function that creates a plot with the formatted table. This would be the fallback for any other output that does not accept HTML or LaTeX (for docx, github_document and others)

Hi, I was wondering whether you had thought about providing a function to allow this to be used in Shiny Apps. So we would need some kinda renderCondformat for the server.R and then a condformatOutput for the ui...

Have you thought about this?

I am using condformat package to color each coloumn of a dataframe. I am able to do single column but don't know how to apply for all columns without using colnames (as they change dynamically)

I used the following code to achieve conditional column coloring for single column

data(iris)
library(condformat)
dat<-iris[c(1:5,70:75, 120:125),]
condformat(dat) %>%
rule_fill_discrete(Sepal.Width,
expression = Sepal.Width > Petal.Length)
The above code changes the color of the column (sepal.width) successfully. But what I want is to apply the same condition for all columns (for example c(1:5) where without knowing the column names. I tried like . and x etc. but nothin

Hi, take a look here: http://stackoverflow.com/questions/41402379/how-can-i-pass-character-strings-as-independent-parameters-after-a

Perhaps the functions being exposed by Gregor and by myself (Amit Kohli) could be useful inclusions to this packet.

The only addition I could think of could be to extend the function to accept inputs for thresolds and colors. Something like:

I'm close with this:

color.picker <- function(z) {
      if (is.na(z)) {                return(0)
      } else if (z < 5) {            return(1)
      } else if (5 >= z & z < 10) {  return(2)
      } else {                       return(3)
      }
    }
    
    CondFormatForInput2 <- function(Table,VectorToColor,VectorFromColor) {
        cf <- condformat(Table)
        input = data.frame(Val=VectorToColor,Comp=VectorFromColor)
        fills2_ = purrr::map2(input$Val,.y = input$Comp,.f = function(x,y) rule_fill_discrete_(x, expression =
                                                                                        sapply(Table[[y]],color.picker),colours = c(`0` = "white", `1` = "red",`2` = "lightyellow", `3` = "lightgreen")))
        res_ = Reduce(f = "+", x = fills2_, init = cf)
        res_
      }
      
      CondFormatForInput2(iris,
                        c("Petal.Width","Petal.Length"),
                        c("Petal.Width","Petal.Width"))

But there's two problems with my approach: 1) the parameter that decides the color is somehow Column 1 (in the xample above it's the Sepal.Length) rather than whatever I'm passing as y 2) This is still messy... I'm passing the sapply portion w/ the colors in the call, not in the color.picker function... which is messy and weird.

I think this would be really cool if the color.picker function accepted a df as input, with column 1 being the thresholds, and column 2 the colors... and then passing that modified color.picker function to the map2 command in CondFormatForInput2.

What do you think?

Hey,

It would be cool if this package were updated to support the recently added shiny async render functions. At the moment, renderCondformat() does not handle promises.

Here is a minimal repro:

library(shiny)
library(future)
library(promises)
library(condformat)
library(dplyr)

ui <- fluidPage(
  actionButton("execute", "Execute"),
  helpText("The synchronous table output"),
  condformatOutput("sync_table"),
  helpText("The asynchronous table output"),
  condformatOutput("async_table")
)

server <- function(input, output) {
  
  our_sync_data <- reactive({
    input$execute
    
    data.frame(type = c(rep("A", 5), rep("B", 5)),
               n = seq(1,100,10))
  })
  
  our_async_data <- reactive({
    input$execute
    
    future({
      
      data.frame(type = c(rep("A", 5), rep("B", 5)),
                 n = seq(1,100,10))
    })
  })
  
  output$sync_table <- renderCondformat({
    
    our_sync_data() %>% 
      condformat() %>% 
      rule_fill_discrete(type, expression = n <= 50,
                         colours = c("TRUE" = "yellowgreen", "FALSE" = "tomato"))
  })
  
  output$async_table <- renderCondformat({
    
    our_async_data() %...>% 
      condformat() %...>% 
      rule_fill_discrete(type, expression = n <= 50,
                         colours = c("TRUE" = "yellowgreen", "FALSE" = "tomato"))
  })
}

shinyApp(ui, server)

And here is the output of sessionInfo():

> sessionInfo()
R version 3.4.2 (2017-09-28)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS High Sierra 10.13.4

Matrix products: default
BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRlapack.dylib

locale:
[1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] promises_1.0.1   dplyr_0.7.4      condformat_0.7.0 future_1.8.1     shiny_1.0.5 

Thank you!

Hi,

I'd like to color different columns in the xlsx that I create, by using several rules for the fill gradient.

Everything looks nice in the R Studio Viewer, but any then the firat rule_fill_gradient seem to be ignored by condformat2excel.

When e.g. doing the following, only column 1 is colored with a green gradient...

output <- condformat(df4) + rule_fill_gradient(1, expression = overall_score, low = "white", high = "green", na.value = "#FFFFFF") + rule_fill_gradient(2, expression = overall_score, low = "white", high = "red", na.value = "#FFFFFF")

condformat2excel(output, output.xlsx)

Current render2excel function only accepts a parameter for workbook name.

Can you add a parameter for sheetname as the current default of sheet1 is not accessible and therefore I am not able to loop through a series of tables that I need to export.

myTable <- condformat(iris[c(1:5, 70:75, 120:125), ]) +
rule_fill_discrete(Species, colours = c("setosa" = "red",
"versicolor" = "blue",
"virginica" = "green")) +
rule_fill_discrete(Sepal.Length, expression=Sepal.Length > 4.6,
colours=c("TRUE"="red"))

condformat2latex(myTable)

Currently there is not a legend system implemented describing the conditional formatting rules.

For instance in:

data(iris)
library(condformat)
condformat(iris) + scale_fill_discrete(Species)

The legend would have three rows, all empty but in the Species column, showing the three possible values with their colour.

data(iris)
library(condformat)
condformat(iris) + scale_fill_discrete(Species, expression = Sepal.Length > 3.5, c("TRUE" = "red"))

The legend here would have two rows. The rows would only be filled in the Species column having as text the expression == TRUE (or FALSE), formatted in red or whatever.

With gradients, showing the extremes would be fine.

I note that when I call condformat2grob the plot is immediately generated and displayed, even if what I'm trying to do is to store the grob in a variable for layout with ggarrange or grid.arrange. In my case, where I have two conditionally formatted tables and one ggplot layed out with ggarrange, this causes knitr to output first the one table, then the second one, and then the composite figure (which itself is just fine).

I seem to have fixed this myself by altering the source code -- I commented out two function calls right at the end of the condformat2grob function: grid::grid.newpage() and grid::grid.draw(gridobj). The subsequent call (invisible(gridobj)) seems to return the grob properly, and my ggarrange layout works as it should.

Knocking out these two lines probably messed up something elsewhere (so not going to make a PR out of this), but for my purposes it works perfectly. Thanks for your work on this -- it's fantastic!

Hi,

Just wanted to say thank you, have been using the condformat package a lot with much success! On the recent release of version 0.9.0 I am encountering the following error when overwriting an existing xlsx workbook.

Warning: Error in openxlsx::saveWorkbook: File already exists!

Any help is much appreciated.

Once 0.7.0 is accepted and binaries are out there, follow http://r-pkgs.had.co.nz/release.html to finish the release steps

Or alternatively without having to explicitly name each column? I'm working on a shiny app and depending on the user selection the table and column names change. Thanks for your time.

I need to test the behaviour for:

• print(condformat(head(iris)))
• knit_print(condformat(head(iris)))
• leaving condformat(head(iris)) at the end of a block

And:

• make sure that the expected output happens. For HTML, condformat2html or condformat2widget. For LaTeX condformat2latex ...
• Make sure it works with html_document, rmd notebooks, pdf_documents, vignettes...

When knitr 1.18 is released update condformat:::guess_output_format leveraging on exported functions: rstudio/rmarkdown#649

cf <- condformat(data[1:nrow(data),]) %>%
  rule_text_color(col1,
  expression = ifelse(col1 < col2 | col1 > col3, "red", ""))
  %>%
  rule_text_color(col4,
  expression = ifelse(col4 < col5 | col4 > col6, "red", ""))

condformat:::condformat2excelsheet(cf, wb, "Sheet1")
saveWorkbook(wb, file = "formatted_file.xlsx", overwrite = TRUE)

This doesn't seem to be working. Would appreciate some help!

Please use parse_quo() instead.

Can condformat handle "invalid" column names?

I tried some using:

myvar = "invalid col name %"
as.name(myvar)

but that leads to a parse error. Can I change colnames after the condformat() statement? What's the best way to do this?

Thanks, Eric.

Is there any way to change from the default of 10 rows per page to some other value in the code (rather than changing the value once the table is produced?)

Hi,

I am the author of the openxlsx package. With the new version some of the test in your package fail.

Could you please check the dependencies?

If you have questions please feel free to contact me.

This is a feature request ;).
It would be nice to be able to control the tabulation (and page size) in shiny as well. I see that there is a parameter tabulate in the print function, however I haven't been able to force this form renderCondformat.

I want to create a rule that applies to several columns and make the expression dependant on the column where it is applied.

Proposed syntax:

    condformat(iris) %>%
      rule_fill_discrete(c(Sepal.Length, Sepal.Width), .col > 3)

Equivalent to:

    condformat(iris) %>%
      rule_fill_discrete(Sepal.Length, Sepal.Length > 3) %>%
      rule_fill_discrete(Sepal.Width, Sepal.Width > 3)

All tests that use suggested packages should check to make sure the package is installed.

(This helps with revdepcheck)

Thank you,
Barret