timelyportfolio

I believe this should now be fixed with 9637a33c113070c1329a24ea2a35908882fb6f2b. Could you try it out?

What you did sounds right. Let me dig a little deeper. Are you getting this with `saveWidget` or with `rmarkdown`?

is there a directory like `test_files` that still exists? I wonder if there is something stuck in cache.

ok, great. I'm not sure what takes it so long. I am guessing all the bootstrap files. Glad it worked. That was going to be my next suggestion :)

not that I am aware of, but they should only be temporary and disappear after save. Is this not the way it works on your machine?

Unfortunately, there is no way around this currently. To add formatting, `formattable` converts the contents of each cell to a `HTML` element. `datatable` does not understand how to filter `HTML`...

We can eliminate the `knitr` fairly easily. See https://github.com/timelyportfolio/sparkline/blob/update/sparkline/inst/examples/sparkline_histogram.R for an ugly start to iterating through this. 

@mexindian, here is my first pass at a second draft for your proposed code: ``` library(dplyr) library(formattable) library(sparkline) formattableWithSL % summarize_at( vars(matches(SLfield)), function(x){ as.character(htmltools::as.tags( sparkline(c(x),...) )) } ) ft %...

As another example to consider, let's try to add sparkline boxplots as a summary line to `mtcars`. Here is my first attempt.  ``` library(dplyr) library(sparkline) library(formattable) library(htmltools) spark_box %...

Two convenience reasons - convert to 1. character with 2. formattable formatting. I use all the time so I will probably try to convince @renkun-ken to export eventually.