Hi all,
long I have been thinking about proper rendering of math in the HTML form
of R documentation. As you know, you can write \eqn{} in your .Rd files and
this is nicely rendered into the PDF Reference manual of the package with
the aid of TeX. However, that is not the case in the aforementioned HTML
version that is used the most in my experience (using RStudio or help
function in your console). I think R is the best language for statisticians
and other data-driven fields, where formal definitions of key concepts are
necessary and widely used in the documentation, unfortunately quite
unusable for more complicated equations.
Recently I have stumbled upon an interesting approach to this issue, see
https://cran.r-project.org/package=mathjaxr, but it seems to me as some
weird kind of monkey patching. All packages should be able to benefit from
proper math rendering without any dependencies, in my opinion. I think it
should not be much of a problem utilizing mathjax or other similar library
to enable that. Note we already know what supposed to be math in .Rd (and
we parse and process it in a special way in the PDF routine), the thing is
to render it, not typeset in italics, as it is the case nowadays.
I would be happy to hear any opinion of yours!
Best,
Jan Netik
[[alternative HTML version deleted]]
On 27/05/2021 2:34 p.m., Jan Net?k wrote:> Hi all, > > long I have been thinking about proper rendering of math in the HTML form > of R documentation. As you know, you can write \eqn{} in your .Rd files and > this is nicely rendered into the PDF Reference manual of the package with > the aid of TeX. However, that is not the case in the aforementioned HTML > version that is used the most in my experience (using RStudio or help > function in your console). I think R is the best language for statisticians > and other data-driven fields, where formal definitions of key concepts are > necessary and widely used in the documentation, unfortunately quite > unusable for more complicated equations. > > Recently I have stumbled upon an interesting approach to this issue, see > https://cran.r-project.org/package=mathjaxr, but it seems to me as some > weird kind of monkey patching. All packages should be able to benefit from > proper math rendering without any dependencies, in my opinion. I think it > should not be much of a problem utilizing mathjax or other similar library > to enable that. Note we already know what supposed to be math in .Rd (and > we parse and process it in a special way in the PDF routine), the thing is > to render it, not typeset in italics, as it is the case nowadays. > > I would be happy to hear any opinion of yours!You are being very rude. The problem has been solved by mathjaxr. If you don't like their approach, don't ask someone else (i.e. R Core) to come up with a different solution, do it yourself. I hope my opinion makes you happy! Duncan Murdoch
Just to clarify the confusion, let me rephrase that. I see this as a request to
add support for converting \eqn{} and friends to MathJax in R html documentation
(both in core packages as well as all contributed packages). That sounds like a
reasonable request to me, but I would not volunteer to implement it so I
don't know how difficult that would be. Obviously, that has to happen in
core R since it is part of the package building process.
The other part of the e-mail was unfortunately just conflating different things
such as alternative ways of creating MathJax output in packages which is already
possible and a solved problem as Duncan pointed out. You can write R
documentation which uses MathJax in your package, but you cannot render existing
R documentation with MathJax (currently, if I understand that correctly). I hope
this clarifies things a bit.
Cheers,
Simon
> On 28/05/2021, at 6:34 AM, Jan Net?k <netikja at gmail.com> wrote:
>
> Hi all,
>
> long I have been thinking about proper rendering of math in the HTML form
> of R documentation. As you know, you can write \eqn{} in your .Rd files and
> this is nicely rendered into the PDF Reference manual of the package with
> the aid of TeX. However, that is not the case in the aforementioned HTML
> version that is used the most in my experience (using RStudio or help
> function in your console). I think R is the best language for statisticians
> and other data-driven fields, where formal definitions of key concepts are
> necessary and widely used in the documentation, unfortunately quite
> unusable for more complicated equations.
>
> Recently I have stumbled upon an interesting approach to this issue, see
> https://cran.r-project.org/package=mathjaxr, but it seems to me as some
> weird kind of monkey patching. All packages should be able to benefit from
> proper math rendering without any dependencies, in my opinion. I think it
> should not be much of a problem utilizing mathjax or other similar library
> to enable that. Note we already know what supposed to be math in .Rd (and
> we parse and process it in a special way in the PDF routine), the thing is
> to render it, not typeset in italics, as it is the case nowadays.
>
> I would be happy to hear any opinion of yours!
>
> Best,
>
> Jan Netik
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>