Fwd: Common Lisp Stat

[A.J. Rossini]

On Sunday, April 14, 2013 9:32:19 PM UTC+2, Stefan Mandl wrote:

On Sun, 14 Apr 2013 09:15:03 -0700, casanova051570 wrote:

> using ccl1.8 64 bit on debian linux via quicklisp install  (ql:quickload "cls") gives error message
>
> Reader error on #<BASIC-FILE-CHARACTER-INPUT-STREAM ("/home/nelson/lib-asdf/quicklisp/dists/quicklisp/software/lisp-matrix-20130312-git/src/unittests/unittests.lisp"/4 ISO-8859-1) #x30200349BC2D>, near position 9021, within "((lambda ,(mapcar #â":
> Comma not inside backquote
>    [Condition of type CCL::SIMPLE-READER-ERROR]
>
> and when i skip this i get other errors related to unittests etc..
>
> any suggestions on how to fix this problem?

Seems to be a file/character encoding problem.
Shortly after position 9021, there's something that looks like a backquote, but actually is a:

             position: 9029 of 11301 (80%), column: 24
            character: ’ (displayed as ’) (codepoint 8217, #o20031, #x2019)
    preferred charset: unicode (Unicode (ISO10646))
code point in charset: 0x2019
               syntax: .         which means: punctuation
             category: .:Base, c:Chinese, h:Korean, j:Japanese
          buffer code: #xE2 #x80 #x99
            file code: #xE2 #x80 #x99 (encoded by coding system utf-8-unix)

Please note that your Lisp seems to think that the file is encoded in ISO-8859-1,
while Emacs thinks it's UTF-8.

This particular case occurs in a comment so -- as a quick
solution -- you probably simply could delete it
(and the other one) a few lines after.


--

Stefan

[Peter Schmiedeskamp]

I had a similar encoding issue with one of the dependencies and SBCL on OS X: https://groups.google.com/d/msg/lisp-stat/CbmMP2LiH3E/uYqg1biOKjEJ

Ultimately, I switched the default encoding and it fixed the problem. Perhaps there is some analogous command in CCL? Oddly, CCL worked fine for me.

-p

[Steven Núñez]

Morning Gents,

It’s been rather quiet around here lately. Anyone still working on lisp-stat? I have need for some basic statistics, but unfortunately the requirement for graphics means that I’m going to have to go back to the original, xlisp-stat and X11. Ugh. X11 just doesn’t feel right when working on a Mac.

As part of this, I’d like to keep as close to common-lisp stat as I can. Does anyone know where to obtain the original common lisp port that Tierney started? The one that, I assume, common-lisp-stat was derived from?

Cheers,

- Steve

[Steven Núñez]

As a follow-up, I found an alpha version: CLS1.0A1 at http://ftp.uni-bayreuth.de/math/statlib/xlispstat/Src/other/CL/. Sadly it doesn’t use CLOS, but the Xlisp object system. Does anyone know if there are any later versions, perhaps with better CLOS support?

Regards,

- SteveN

--
You received this message because you are subscribed to the Google Groups "Common Lisp Statistics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lisp-stat+...@googlegroups.com.
To post to this group, send email to lisp...@googlegroups.com.
Visit this group at http://groups.google.com/group/lisp-stat.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[A.J. Rossini]

You can simply roll back the git repo to day one, and that is Luke's version that you cite'd.

If you move forward on that, you can see where I've patched up the CL version of Luke's version of the lisp-stat prototype object system, and got the C code loaded with an older version of CFFI.

Yes, I'm still working on it, I've got forgotten to commit a big whopping load of changes!

I need to do this to clean up the ASDF file for ASDF 3, Xach pointed out some challenges, and I need to upload the mods.

--
best,
-tony

blind...@gmail.com
Muttenz, Switzerland.
"Commit early,commit often, and commit in a repository from which we can easily roll-back your mistakes" (AJR, 4Jan05).

Drink Coffee:  Do stupid things faster with more energy!

[Steven Núñez]

Thanks. Interestingly, I read Luke’s “Recent Developments and Future Directions in Lisp-Stat” on the train today and noticed that many of the features mentioned there are implemented (or being implemented) in CLS. I’ve got to say though that the importance of CLS being fully quicklisp-able can’t be understated. I haven’t tried to set-up CLS for a while now, but my last attempt was a 3 day effort that ended nowhere. I set up Xlisp-stat, along with coffee, breakfast and email, before 1:00pm and have graphics working. If a community is going to develop around CLS, it’s got to have a lower barrier to entry than it does now.

Now that I’ve got xlisp-stat work I can breath a bit easier and might give CLS a whirl again. I need to take a very pragmatic approach (I have a problem to solve) and wouldn’t mind grafting on AquaTerm, if it’s feasible to do so. What’s the current ‘state of CLS’ as far as CLOS goes? Is this going to be a vertical ascent or a less than 45 degree learning curve to implement AquaTerm graphics? I don’t need to worry about compatibility, as long as it runs on a Mac and gets the job done that’s good enough for now.

Cheers,

- SteveN

[David Hodge]

Hey Steve,

How is Thailand? Still there?

You are right, things have been pretty quiet for a while - a variety of life changing events recently coupled with a punishing travel schedule has left little time for lisp.

However, after spending a week or so slogging through some Java  stuff, i am absolutely itching to get back to sanity so i want to start up soon.

Hopefully, Tony will make the required patches for CLS and we can start working - Tony, need help with that?

A program of work for me is (in no particular order)

1. Data tables. Mirko V created a nice implementation based on PCL and I think thats the best candidate. The current datatable implementation in CLS really does not add much value in my opinion. Some thought would have to be given to datatable->grid/lisp-matrix

2. Graphics. To date, the default has been to just use gnuplot. The reasoning is that its easy, there are a plethora of interfaces available and stable across platform. However, i think it is nice if we can have a 'batteries included" approach. To that end there is an SDL based simple plotting package that can be used for this. it just means that for simple things (eg, following the original lisp stat tutorial ) you can just get going without any installation or integration issues. However, its important that if people  want to use a different graphics package then they are able to. One thing i have experimented with is an interface to the javascript world, so its easy to drive a js package like D3 or flot.

This approach has downsides of course, in that it might not be easy to share things across different implementations, but i tend to regard CLS are more a toolkit than a package, so it can be the users choice to use something specific to their needs and CLS provides the numerical/statistical substrate supporting that.

The long long term objective would be to have the inbuilt graphics implement grammar of graphics of ocurse.

3. regression models. The lisp-stat object model is kind of redundant in comparision to CLOS. I don't see much point in having it around, but we need to assess how to "closify" that area. I have absolutely no idea yet where that might lead

4. Data  - at the  moment my tiny (stalled)  project its to build a NETCDF interface, in CL as many of my data sets of interest are distributed in that format. When I get back from Boston i am aiming to finish it.

5. time series analysis. AR models, fourier analysis and such. A question here might be whether we leverage R for some of this initially and incrementally implemented. I have some AR models in lisp, but testing and validation is an issue.

6. housekeeping. For me CLS installs without issue but thats after much work. I would hope that once Tony commits his changes we would have a cleanly installable baseline. The other thing is to review the test suite and make sure everything is fully understood in terms of coverage and why tests are failing.

that will be enough for now I think!

[A.J. Rossini]

To confirm, David's list is pretty accurate.    We need a means to manipulate data, then describe and view, and the infer about it.

David's list covers that!