[R] Wolfram Engine for Developers

J C Nash pro|jcn@@h @end|ng |rom gm@||@com
Sat Jul 6 14:46:03 CEST 2019


Nobody has mentioned Julia. Last year Changcheng Li did a Google Summer of Code project to
add automatic differentiation capability to R. autodiffR package was result, but it is still
"beta". The main awkwardness, as I would guess for Wolfram and other wrappings, is the
non-R side having "updates" so the interface changes slightly but enough to force changes
in the wrapper.

As Jeff points out, those with itch must be the scratchers.

JN

On 2019-07-06 7:15 a.m., Jeff Newmiller wrote:
> I don't get it. I have been quite content to use Maxima when I want symbolic manipulation for many years. I have not felt this pressure you imply that R should do everything, and I have tried both Mathematica and Wolfram Alpha at times and found neither of them so compelling that I felt the slightest bit jealous of people who have licensed access to Wolfram's tools and services.
> 
> This satisfaction with an available separate tool may arise from my expectation that symbolic manipulation should occur in the course of developing theoretical foundations for numerical work as a separate stage from the numerical work itself. Having such separation fits the pattern of writing papers and documentation for human consumption separately from implementing algorithms... which isn't the only way to do things but has its benefits. (I have not even felt any incentive to use Ryacas package to interface with Yacas. People looking for some integration with R may find Ryacas fills some of that need, but it is not as feature rich as Maxima. [1])
> 
> Of course, none of that proves that there shouldn't be some such capability to interact with WED services... but as always the onus for implementation belongs to those with the itch, and it does look to me like license compatibility will be a constraint. Specifically, R+WED would inevitably carry restrictions that R+Ryacas or Python+SymPy will not.
> 
> FWIW I am just a satisfied user of R (and Maxima), and speak only for myself.
> 
> [1] https://en.m.wikipedia.org/wiki/Comparison_of_computer_algebra_systems
> 
> On July 5, 2019 11:41:41 PM PDT, Anarcocapitalista Socialdemocrata <anarco.socialdemocrata using gmail.com> wrote:
>> I think the Wolfram license is for non-profit, i.e. users and NGOs.
>> Companies making a profit should buy a license. IMHO, the strategy is
>> to
>> increase Wolfram user base.
>>
>> For R and RStudio, it would make sense to add Wolfram to become a
>> comprehensive solution to do science. R is of course mostly
>> statistical. R
>> competes very well with Python in the areas R does well. But in
>> anything
>> not numerical, but symbolic, R lacks a lot. Instead, Python has SymPy
>> and
>> Sagemath.
>>
>> So, a scientist may switch from R to Python, due to the combined
>> numerical
>> and symbolic capabilities.
>>
>> Instead, if RStudio could use Wolfram, then a scientist would have
>> better
>> symbolic than Python (or the same, since Jupyter already interfaces
>> Wolfram), RStudio which is better than Jupyter, better statistics than
>> Python and better c++ integration than Python.
>>
>> At least me, I am in this conundrum. With integration to Wolfram, it
>> would
>> be a no-brainer to remain with R.
>>
>> On Fri, Jul 5, 2019, 20:28 peter dalgaard <pdalgd using gmail.com> wrote:
>>
>>> Even from a cursory glance, it is clear that this product is not Free
>>> Software, not even Free as in Beer. It cannot be used by end users of
>> open
>>> source projects unless you apply for a special license from Wolfram,
>> which
>>> I strongly suspect would render the whole project in violation of the
>> GPL
>>> license that R has.
>>>
>>> It is possible that you could write an interface from R to WED. The
>>> licensing questions around "derived works" are a bit murky, but as
>> far as I
>>> know it is OK for a GPL'ed software to _use_ a commercial software,
>>> assuming that the user has the appropriate license. However, I fail
>> to see
>>> that such an interface would be a major selling point for R.
>>>
>>> -pd
>>>
>>>> On 5 Jul 2019, at 19:30 , Jordi Molins
>> <anarco.socialdemocrata using gmail.com>
>>> wrote:
>>>>
>>>> Wolfram Engine for Developers is now free (under some
>> circumstances). Is
>>> it
>>>> possible to call Wolfram from R, especially from RStudio?
>>>>
>>>> Being able to do this would significantly increase the potential of
>> R, I
>>>> believe.
>>>>
>>>>       [[alternative HTML version deleted]]
>>>>
>>>> ______________________________________________
>>>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>>> PLEASE do read the posting guide
>>> http://www.R-project.org/posting-guide.html
>>>> and provide commented, minimal, self-contained, reproducible code.
>>>
>>> --
>>> Peter Dalgaard, Professor,
>>> Center for Statistics, Copenhagen Business School
>>> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
>>> Phone: (+45)38153501
>>> Office: A 4.23
>>> Email: pd.mes using cbs.dk  Priv: PDalgd using gmail.com
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>
>> 	[[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>



More information about the R-help mailing list