[Pdl-devel] Problem with Dev.pm and inline

On the current git, inline pdlpp code is failing to compile. It appears to be a problem with the "whereami" sub, which seems to be trying to find some files in Basic/Core in the installation/build tree of which the current working directory is a part. To reproduce: - CD to some direc

Re: [Pdl-devel] fitpoly1d()

I made that change during the whole g77-to-gfortran switch several years ago, when it was hard to get slatec to compile on many platforms. I have no objection to switching it back. MatrixOps was an attempt to put at least *something* in place that would compile whenever the core of PDL would —

Re: [Pdl-devel] PDL master failing tests on Travis

I submitted a patch last night to fix Inline::Pdlpp. It apparently broke the build on some systems (but not mine). I’ll back it out and stick it in a branch. (Zakariyya has already brought it to my attention). > On Apr 9, 2015, at 10:12 AM, Ed wrote: > > Dear PDL devs, > > I just ran a q

Re: [Pdl-devel] PDL master failing tests on Travis

(that’s “from 1 to 0”) > On Apr 9, 2015, at 10:47 AM, Craig DeForest wrote: > > OK, I’ve backed out the commits from last night. There were three functional > changes. Two were in Dev.pm: > - changing whereami() to whereami_any() in pdlpp_stdargs_int (hopefully > harmles

Re: [Pdl-devel] PDL master failing tests on Travis

eed of regularization. It’s also hard to test because different systems work differently. (Last night’s commits tested 100% OK on my MacOS system…) > On Apr 9, 2015, at 10:29 AM, Craig DeForest wrote: > > I submitted a patch last night to fix Inline::Pdlpp. It apparently broke the > bui

Re: [Pdl-devel] Comms with Christian Soeller?

Ed, the last address I have for Christian is "c.soel...@auckland.ac.nz ”. (Hi, Christian!) > On Apr 10, 2015, at 7:17 PM, Ed wrote: > > Dear PDL devs, > > I am trying to email Christian to ask him to delete 3 old versions of PDL > from PAUSE so that “cpanm -

Re: [Pdl-devel] 2 slices.pd questions

That "$a = " appears to be my problem, since it's my commit. It is "harmless". I would've expected it to be nailed by "use strict" - but I don't think the environment in which the pp_def call is running, is using strict. The "IV" declaration should probably be changed to a "STRLEN". The "len"

Re: [Pdl-devel] fixing sf.net bug #378

I do not think this would be difficult to fix. I will try to reproduce and isolate it tonight. > On Apr 22, 2015, at 8:33 AM, Chris Marshall wrote: > > We're down to the last stages of the PDL-2.008 release but > a recent bug report shows a significant regression since > PDL-2.006 (i.e., both

[Pdl-devel] PDL book -- upload to ArXiV?

Hey, everyone, It seems to me that putting the PDL book in the ArXiV might be a nice way to get it “out there” and indexed (and therefore citable). It is well within the bailiwick of the Computer Science branch. Any thoughts? --

Re: [Pdl-devel] PDL::FFTW deletion

I’m not necessarily advocating, but have you considered instead truncating it to a documentation file with a reference to FFTW3? > On Jun 8, 2015, at 10:42 AM, Chris Marshall wrote: > > I'm planning to remove the PDL::FFTW distribution from CPAN. Does anyone > have any suggestions or comment

Re: [Pdl-devel] Faster PDL Development Cycle---But How?

I apologize to everyone for staying silent so long. I support the plan for a dual pathway. I have been of two minds about the current resurgence: on the one hand I am delighted that at the work Aki and Ed are doing; on the other hand I have been worried for stability of existing projects of som

Re: [Pdl-devel] Faster PDL Development Cycle---But How?

Zaki, not Aki. Darned autocorrect... (Mobile) > On Aug 31, 2015, at 4:51 AM, Craig DeForest wrote: > > I apologize to everyone for staying silent so long. I support the plan for a > dual pathway. I have been of two minds about the current resurgence: on the > one hand I am

Re: [Pdl-devel] SciPDL-2.013 anyone?

Looks like a good place to use the $T macro instead. That guy DeForest looks like kind of a hack… > On Sep 3, 2015, at 10:19 AM, Karl Glazebrook wrote: > > Hi Chris > > > It’s this cheesy code in ufunc.pd > > pp_def( > 'average', > HandleBad => 1, > Pars => 'a(n); int+ [o

[Pdl-devel] Problems with current master in git?

I just cloned the master branch from "https://github.com/PDLPorters/pdl.git”, and it won’t compile on MacOS — several argument type errors kill the compilation. Is this a known issue? Should I still be using sourceforget instead? ---

Re: [Pdl-devel] Problems with current master in git?

erated. >make[2]: *** [pdlsections.o] Error 1 >make[1]: *** [subdirs] Error 2 >make: *** [subdirs] Error 2 I have no idea what “PDL_Anyval” is, I don’t recall it being in the codebase before… > On Sep 28, 2015, at 11:49 AM, Zakariyya Mughal wrote: > > On 2015-09-28 at 11

Re: [Pdl-devel] Problems with current master in git?

issue-specific branch? > On Sep 28, 2015, at 3:29 PM, Craig DeForest wrote: > > So my issue is I clone the repository, check the dependencies (all OK), say > “perl Makefile.PL; make”, and end up with: > > > > >> cc -c "-I/usr/local/src/pdl/Basic/Core

Re: [Pdl-devel] integer exponentiation support for **

Sorry I missed the first iteration. You want something like the $T macro, which will substitute source code snippets depending on the $GENERIC type. But by the time you see the PDLs they will have been coerced to a common type, in a vanilla PP routine. I suggest passing the second argument in

Re: [Pdl-devel] integer exponentiation support for **

Yes, and that can be a pain. But it is not so bad to roll one's own for a corner case like that. (Mobile) > On Oct 3, 2015, at 3:14 PM, Chris Marshall wrote: > > Don't you lose the PP threading for the OtherPar PDL *? > >> On 10/3/2015 17:11, Craig DeForest

[Pdl-devel] PDL::Graphics::Gnuplot update - requests/gripes?

I’m planning a PDL::Graphics::Gnuplot release to coincide with Chris’s awesome PDL release next Monday. This is a good time for anyone to gripe and/or make a (smallish) feature request before I zip up the current git in a couple of days and send it to PAUSE. The main functional update for P::G

Re: [Pdl-devel] link

That *is* pretty cool! I didn’t realize there were so many papers indexed that mention PDL. Nice. > On Nov 10, 2015, at 4:39 PM, Karl Glazebrook wrote: > > can someone include this link on the PDL home page? > > tinyurl.com/mst645o > > It is pretty cool! > > Karl > > > --

Re: [Pdl-devel] PDL atan defaults to float

Yes, I believe this is a Bug in the conversion code logic. Currently we use the smallest possible type that can contain a value. Values that fit in a float get promoted to float and those that require a double get promoted to double. Here’s a four-atan trial, showing the autopromotion using a

Re: [Pdl-devel] PDL atan defaults to float

; or other specific type, smallest size, or ... > > --Chris > > On Wed, Nov 18, 2015 at 4:58 PM, Craig DeForest <mailto:defor...@boulder.swri.edu>> wrote: > Yes, I believe this is a Bug in the conversion code logic. Currently we use > the smallest possible type that can

Re: [Pdl-devel] PDL support in perl6?

Perl 6 has some syntactic features that were introduced specifically to support us. I have not looked into 6 recently, beyond noting that there is an official release. > On Mar 1, 2016, at 2:11 PM, dhunt wrote: > > Hi all: I'm beginning to look into introducing perl 6 into our > production

Re: [Pdl-devel] Fwd: Constructing multi-level perl list from a PDL

“dog” does it one dimension at a time, so if you’ve got a 2D PDL, you can do: @nested = map { [ $_->dog ] } ($im->dog); > On Apr 14, 2016, at 1:04 PM, dhunt wrote: > > > Hi all: I've got what might seem like a silly question: Is there any > PDL routine which will > convert an N-dime

Re: [Pdl-devel] PDL magic is not PERL magic?

We do have a parallel implementation of something magic-like: hooks to kick off a process on evaluation of a PDL. The "PDL magic" hooks, IIRC, are to kick off lazy dataflow — e.g. “EquivCPOffsets” blocks in PP, or anything involving the internal “trans" structure. I am pretty sure they’re not

Re: [Pdl-devel] [Pdl-general] PDL-2.016 released

t; that emerged since PDL-2.015 last November. > > It would not have happened without the work > of some diligent PDL developers and users with > special acknowledgement to contributions from > Derek Lamb, Craig DeForest, and kmx. > > Enjoy! > The PDL Development Tea

Re: [Pdl-devel] 20th Anniversary OMG

I’m definitely FOR a new publication outlining the various bits of awesomeness that have come in under Chris’s direction. Perhaps we could collect a half-dozen to a dozen science highlights as was done for the web page way back when we had the big flag day. > On Jun 8, 2016, at 2:14 AM, Chri

[Pdl-devel] Question on the constructor

So David’s recent “ping” on the long-standing desire to propagate bad values in the constructor got me thinking… Right now the constructor is totally badval-ignorant. In fact, it uses the global variable $PDL::undefval to pad PDLs, and if $PDL::undefval isn’t set, it makes up a reasonable choi

Re: [Pdl-devel] sourceforge

Hear, hear! I’m all for it. > On Jun 12, 2016, at 8:41 PM, Karl Glazebrook wrote: > > The sourceforge web set looks more and more like a ‘rent by the hour’ cheap > hotel every time I go to it. Also there are some dodgy practices going on > there which I don’t think we want to be associated wi

Re: [Pdl-devel] Question on the constructor

Hearing crickets, I went ahead and implemented the hybrid scheme, in the “badval-constructor” branch. I’m waiting for David to give it a spin, then I’ll merge into master. > On Jun 12, 2016, at 7:56 PM, Craig DeForest wrote: > > So David’s recent “ping” on the long-standing

Re: [Pdl-devel] Question on the constructor

a merge, since it doesn't touch too much. > > David > > On Mon, Jun 13, 2016 at 11:23 AM, Craig DeForest <mailto:defor...@boulder.swri.edu>> wrote: > > Hearing crickets, I went ahead and implemented the hybrid scheme, in the > “badval-constructor” branch. I’m

Re: [Pdl-devel] Question on the constructor

gt; is not passing test t/constructor.t when BADVAL_USENAN and/or BADVAL_PER_PDL > is true: see https://travis-ci.org/PDLPorters/pdl/builds/137405085. Might be > worth taking a look. > > Derek > >> On Jun 13, 2016, at 8:32 PM, Craig DeForest >> wrote: >> >> Thanks

Re: [Pdl-devel] [pdl:bugs] #427 PDL->null->copy returns an empty (not null) piddle

Copying a null PDL implicitly reads data from the null PDL (to copy it), which is invalid. The behavior is certainly allowed based on that. That said, I think it would be more consistent and less surprising to either: - retain nullness across copying and severing. - throw an erro

[Pdl-devel] PDL::Transform::Color

Is there a general desire for this subclass of Transform? I’m coding it up for some other projects. It is intended to be, for color representations, sort of what PDL::Transform::Cartography is to map projections. At the moment it’s living in my localpdl github tree, but I’d be happy to either

[Pdl-devel] PDL::Transform::Color module on github

Hi, all, I’ve put enough stuff into PDL::Transform::Color that it needed unpacking from its autoload hack. I’ve set up a github repository for it at "https://github.com/drzowie/PDL-Transform-Color”, and it looks nearly ready for a CPAN release. At the moment I’m holding on feedback from Dav

Re: [Pdl-devel] Shuffling piddle by last dimension

To shuffle in first dim: $b=$a->( qsorti(random($a->dim(0; Stupid index tricks left as an exercise... :-) (Mobile) > On Mar 5, 2017, at 4:56 PM, Sergey Kolychev > wrote: > > Hello all, > Hope you do not mind a quick question (I did search through the docs before > :-)) > Currently I am

[Pdl-devel] PDL::Graphics::Gnuplot and PDL::Transform::Color: new color treatments available

PDL::Transform::Color now has support for pseudocolor mapping, and a basic suite of named color tables. They are accessible via the “t_pc” (“pseudocolor”) and “t_pcp” (pseudocolor, perceptual) transforms. The difference between t_pc and t_pcp is how luminance is treated. t_pc is for handlin

Re: [Pdl-devel] API change for xvals, yvals, zvals...

I’m for it. > On Apr 30, 2017, at 1:01 PM, Chris Marshall wrote: > > Per sf.net feature request #82: > > https://sourceforge.net/p/pdl/feature-requests/82/ > > Now that we have the PDL_Indx data type for > indexing operations, it would seem to make > sense that operations returning index val

Re: [Pdl-devel] 'cat' output datatype based on first input

Funny, I hadn’t noticed — I use pdl(@list_of_piddles) in all cases, since it’s faster and does the Right Thing. > On Jul 17, 2017, at 11:54 AM, Derek Lamb wrote: > > >> On Jul 17, 2017, at 11:13 AM, Derek Lamb > > wrote: >> >> That way cat(@list_of_piddles)->do