Re: [racket-users] Mann-Whitney test?

[Mitchell Wand]

Yo!  This conversation has gone seriously off-topic.  Can y'all carry it on
with some other subject line?
--Mitch

On Tue, Jan 24, 2017 at 11:38 AM, Tim Chase  wrote:

> Here is an interesting alternative: https://www.opencpu.org.
>
> You communicate with R processes via JSON. So, you wouldn't want a lot of
> back-and-forth, but I can see doing some setup logic and data prep in
> Racket, then sending off the JSON request, then interpreting the results.
> OpenCPU happens to be R in the middle right now, but as far as Racket is
> concerned, it is a black box that takes in JSON and returns JSON.
>
>
> On Monday, January 23, 2017 at 1:23:01 PM UTC-5, Leif Andersen wrote:
> > You _could_ use the FFI, but you'd have to use the Racket FFI to go to
> C, and then use R's FFI to go to C, and then coerce all of R's data to
> Racket and all of Racket's data back to R.
> >
> > Because of this (provided no one makes a #lang R first ;)  ), I think it
> would be easier to just spawn up a new R process using something like
> system* or process, and use pipes to send data back and forth directly.
> Yes, all of the data would need to be serialized into streams, but you
> would do an analogous process going from R to C to Racket.
> >
> >
> >
> >
> >
> >
> >
> > ~Leif Andersen
> >
> >
> > On Sun, Jan 22, 2017 at 12:19 AM, James  wrote:
> >
> >
> > > Does anybody have an implementation of the Mann-Whitney U test (or the
> equivalent Wilcoxson Rank test) in Racket?  I imagine I could easily hack
> up my own, but somebody else may have done a better job of it already.
> >
> >
> >
> > I see that no one has answered this question yet.  This is something
> which I would do and have done in R.  So this brings me to a question I was
> going to ask anyway which is how to do multi-language development including
> Racket and R.  Would it be FFI from Racket to R?  I already have a body of
> R code of my own which I would like to be able to call from Racket.  The
> general idea is that It would be an application where things like the GUI
> and networking elements are done in Racket but computations are done in R.
> >
> >
> >
> > James
> >
> >
> >
> >
> >
> > --
> >
> > You received this message because you are subscribed to the Google
> Groups "Racket Users" group.
> >
> > To unsubscribe from this group and stop receiving emails from it, send
> an email to racket-users...@googlegroups.com.
> >
> > For more options, visit https://groups.google.com/d/optout.
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Mann-Whitney test?

[Tim Chase]

Here is an interesting alternative: https://www.opencpu.org. 

You communicate with R processes via JSON. So, you wouldn't want a lot of 
back-and-forth, but I can see doing some setup logic and data prep in Racket, 
then sending off the JSON request, then interpreting the results. OpenCPU 
happens to be R in the middle right now, but as far as Racket is concerned, it 
is a black box that takes in JSON and returns JSON. 


On Monday, January 23, 2017 at 1:23:01 PM UTC-5, Leif Andersen wrote:
> You _could_ use the FFI, but you'd have to use the Racket FFI to go to C, and 
> then use R's FFI to go to C, and then coerce all of R's data to Racket and 
> all of Racket's data back to R.
> 
> Because of this (provided no one makes a #lang R first ;)  ), I think it 
> would be easier to just spawn up a new R process using something like system* 
> or process, and use pipes to send data back and forth directly. Yes, all of 
> the data would need to be serialized into streams, but you would do an 
> analogous process going from R to C to Racket.
> 
> 
> 
> 
> 
> 
> 
> ~Leif Andersen
> 
> 
> On Sun, Jan 22, 2017 at 12:19 AM, James  wrote:
> 
> 
> > Does anybody have an implementation of the Mann-Whitney U test (or the 
> > equivalent Wilcoxson Rank test) in Racket?  I imagine I could easily hack 
> > up my own, but somebody else may have done a better job of it already.
> 
> 
> 
> I see that no one has answered this question yet.  This is something which I 
> would do and have done in R.  So this brings me to a question I was going to 
> ask anyway which is how to do multi-language development including Racket and 
> R.  Would it be FFI from Racket to R?  I already have a body of R code of my 
> own which I would like to be able to call from Racket.  The general idea is 
> that It would be an application where things like the GUI and networking 
> elements are done in Racket but computations are done in R.
> 
> 
> 
> James
> 
> 
> 
> 
> 
> --
> 
> You received this message because you are subscribed to the Google Groups 
> "Racket Users" group.
> 
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to racket-users...@googlegroups.com.
> 
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Mann-Whitney test?

[Leif Andersen]

You _could_ use the FFI, but you'd have to use the Racket FFI to go to C,
and then use R's FFI to go to C, and then coerce all of R's data to Racket
and all of Racket's data back to R.

Because of this (provided no one makes a #lang R first ;)  ), I think it
would be easier to just spawn up a new R process using something like
system* or process, and use pipes to send data back and forth directly.
Yes, all of the data would need to be serialized into streams, but you
would do an analogous process going from R to C to Racket.


~Leif Andersen

On Sun, Jan 22, 2017 at 12:19 AM, James  wrote:

>
> > Does anybody have an implementation of the Mann-Whitney U test (or the
> equivalent Wilcoxson Rank test) in Racket?  I imagine I could easily hack
> up my own, but somebody else may have done a better job of it already.
>
> I see that no one has answered this question yet.  This is something which
> I would do and have done in R.  So this brings me to a question I was going
> to ask anyway which is how to do multi-language development including
> Racket and R.  Would it be FFI from Racket to R?  I already have a body of
> R code of my own which I would like to be able to call from Racket.  The
> general idea is that It would be an application where things like the GUI
> and networking elements are done in Racket but computations are done in R.
>
> James
>
> --
> You received this message because you are subscribed to the Google Groups
> "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to racket-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] Mann-Whitney test?

[James]

> Does anybody have an implementation of the Mann-Whitney U test (or the 
> equivalent Wilcoxson Rank test) in Racket?  I imagine I could easily hack up 
> my own, but somebody else may have done a better job of it already.

I see that no one has answered this question yet.  This is something which I 
would do and have done in R.  So this brings me to a question I was going to 
ask anyway which is how to do multi-language development including Racket and 
R.  Would it be FFI from Racket to R?  I already have a body of R code of my 
own which I would like to be able to call from Racket.  The general idea is 
that It would be an application where things like the GUI and networking 
elements are done in Racket but computations are done in R.  

James 

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[racket-users] Mann-Whitney test?

[Mitchell Wand]

Does anybody have an implementation of the Mann-Whitney U test (or the
equivalent Wilcoxson Rank test) in Racket?  I imagine I could easily hack
up my own, but somebody else may have done a better job of it already.

--Mitch

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.