There's ORLite::PDL http://search.cpan.org/~adamk/ORLite-PDL-0.02/lib/ORLite/PDL.pm
Not DBI, but along the same line of thought. Yes diag sounds like a great tool here. Thanks for pointing it out! Best, Maggie On Thu, Dec 22, 2011 at 10:51 PM, chm <[email protected]> wrote: > On 12/22/2011 10:43 PM, Puneet Kishor wrote: > >> >> On Dec 22, 2011, at 9:29 PM, Maggie X wrote: >> >> Thank you for trying it out. Glad it worked! But I'm not sure why >>> just printing out the numbers would make it work. Was there >>> anything else that you have tried to make it work? >>> >>> >> >> No... just added the `print` statements. That was it. Maybe the >> `print` output resolved to `true` and that was it. You might want to >> alter that test to account for such things. >> >> Many thanks for your help. Now I can have a lot of fun wasting time >> trying to redo stuff in PDL::Stats that I can already do in >> Postgres/R ;-) >> >> >> On a tangential note -- seriously, it would be a great boon if PDL >> could ingest results of DBI queries as piddles, kinda like >> (theoretical code ahead) >> >> $sth->execute; my $piddle = $sth->fetchrow_arrayref; my $result = >> $piddle->do_magic; return $result->to_json; >> > > Two things: it appears that your email app has problems with > correctly encoding line endings (it seems to be a mac only > issue), and I seem to recall a module for reading database > info as piddles. Try checking the ones in the list of > "other" PDL modules I posted in the past week. Sorry, I > can't recall and I'm not using DBI stuff and the moment > myself.... > > Cheers, > Chris > > > piano, piano... >> >> >> >> >> Best, Maggie >>> >>> On Thu, Dec 22, 2011 at 10:13 PM, Puneet >>> Kishor<[email protected]> wrote: >>> >>> Here ya go... I did as you said, and now it says all tests >>>> passed. My guess is that the darn thing would work well if I just >>>> installed it without testing it, but `cpanm` stops as soon as the >>>> first test croaks. >>>> >>>> punkish@mumbai ~/Projects/PDL-Stats-0.5.5$**make test >>>> PERL_DL_NONLAZY=1 /opt/local/bin/perl "-MExtUtils::Command::MM" >>>> "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t >>>> t/stats_basic.t ..... ok t/stats_glm.t ....... ok >>>> t/stats_kmeans.t .... 1/18 R2 0.786191536748329 centroid [ [ >>>> 1.5 5.5] [ 8.5 12] [15.5 19] ] >>>> >>>> cluster [ [1 1 1 1 0 0 0] [0 0 0 0 1 1 1] ] >>>> >>>> ms [ [ 1.25 0.25] [ 1.25 0.66666667] [ 1.25 >>>> 0.66666667] ] >>>> >>>> n [4 3] -6.66666570836583e-08R2 0.786191536748329 centroid >>>> [ [ 1.5 5.5] [ 8.5 12] [15.5 19] ] >>>> >>>> cluster [ [1 1 1 1 0 0 0] [0 0 0 0 1 1 1] ] >>>> >>>> ms [ [ 1.25 0.25] [ 1.25 0.66666667] [ 1.25 >>>> 0.66666667] ] >>>> >>>> n [4 3] t/stats_kmeans.t .... ok t/stats_ols_rptd.t .. ok >>>> t/stats_ts.t ........ ok All tests successful. Files=5, >>>> Tests=147, 1 wallclock secs ( 0.03 usr 0.01 sys + 0.93 cusr >>>> 0.03 csys = 1.00 CPU) Result: PASS No tests defined for >>>> PDL::Stats::Basic extension. PERL_DL_NONLAZY=1 >>>> /opt/local/bin/perl "-MExtUtils::Command::MM" >>>> >>> >> "-e" "test_harness(0, '../blib/lib', '../blib/arch')" t/*.t >>>> t/stats_distr.t .. ok All tests successful. Files=1, Tests=43, 0 >>>> wallclock secs ( 0.02 usr 0.00 sys + 0.13 cusr 0.00 csys = >>>> 0.15 CPU) Result: PASS No tests defined for PDL::Stats::GLM >>>> extension. No tests defined for PDL::Stats::Kmeans extension. >>>> PERL_DL_NONLAZY=1 /opt/local/bin/perl "-MExtUtils::Command::MM" >>>> >>> >> "-e" "test_harness(0, '../blib/lib', '../blib/arch')" t/*.t >>>> t/cdf.t .. ok All tests successful. Files=1, Tests=4, 0 >>>> wallclock secs ( 0.02 usr 0.00 sys + 0.06 cusr 0.00 csys = >>>> 0.08 CPU) Result: PASS No tests defined for PDL::Stats::TS >>>> extension. punkish@mumbai ~/Projects/PDL-Stats-0.5.5$ >>>> >>>> On Dec 22, 2011, at 9:08 PM, Maggie X wrote: >>>> >>>> I've seen cases where the test fails simply because of the >>>>> precision >>>>> >>>> differ >>>> >>>>> on different platforms. That's why I wanted to see what you get >>>>> when you >>>>> >>>> do >>>> >>>>> print $m{ms}->sumover(). But I guess make test doesn't actually >>>>> print out the values to stdout... How about this? >>>>> >>>>> Add >>>>> >>>>> print STDERR t_kmeans_bad(); >>>>> >>>>> 168> is(tapprox( t_kmeans_bad(), 0 ), 1); 169> sub >>>>> t_kmeans_bad { 170> my $data = sequence 7, 3; 171> >>>>> $data = $data->setbadat(4,0); 172> my %m = >>>>> $data->kmeans({NCLUS=>2, NTRY=>10, V=>0}); >>>>> >>>> >> >>>>> print STDERR "$_\t$m{$_}\n" for (sort keys %m); >>>>> >>>>> 173> return sum( $m{ms}->sumover - pdl qw( 1.5 1.9166667 >>>>> 1.9166667 >>>>> >>>> ) ); >>>> >>>>> 174> } >>>>> >>>>> >>>>> Best, Maggie >>>>> >>>>> >>>>> On Thu, Dec 22, 2011 at 9:06 PM, Puneet >>>>> Kishor<[email protected]> >>>>> >>>> wrote: >>>> >>>>> >>>>> I should have added my congratulations and gratitude for >>>>>> taking 2.4.9 >>>>>> >>>>> >> to a >>>> >>>>> point that it builds and installs beautifully, without any >>>>>> heartburn. >>>>>> >>>>> >> Perhaps because most of the dependencies got installed when I >>>>>> installed 2.4.6 via MacPorts. >>>>>> >>>>>> >>>>>> I am doing a lot of stuff using R, so am curious to try out >>>>>> PDL::Stats. For now, I am using R compiled inside Postgres >>>>>> using PL/R... what a wonderful experience to get my data out >>>>>> of Postgres and analyze it right within the database. >>>>>> >>>>>> Wrt PDL::Stats, my guess is the test failure has nothing to >>>>>> do with it being on a Mac/10.7. Perl 5.14.1 should be Perl >>>>>> 5.14.1 no matter what >>>>>> >>>>> >> platform, no? >>>>>> >>>>>> On Dec 22, 2011, at 7:50 PM, chm wrote: >>>>>> >>>>>> On 12/22/2011 7:48 PM, Puneet Kishor wrote: >>>>>>> >>>>>>>> Inspired by the lovely blurb below, I tried to install >>>>>>>> PDL::Stats but >>>>>>>> >>>>>>> failed. Here is the relevant bit from the log of my failed >>>>>> attempt -- >>>>>> >>>>> >> >>>>>>>> ---- PERL_DL_NONLAZY=1 /opt/local/bin/perl >>>>>>>> "-MExtUtils::Command::MM" "-e" >>>>>>>> >>>>>>> "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t >>>>>> >>>>>>> t/stats_basic.t ..... ok t/stats_glm.t ....... ok >>>>>>>> >>>>>>>> # Failed test at t/stats_kmeans.t line 168. # >>>>>>>> got: '' # expected: '1' # Looks like you failed 1 >>>>>>>> test of 18. t/stats_kmeans.t .... Dubious, test returned >>>>>>>> 1 (wstat 256, 0x100) Failed 1/18 subtests >>>>>>>> t/stats_ols_rptd.t .. ok t/stats_ts.t ........ ok >>>>>>>> >>>>>>>> Test Summary Report ------------------- t/stats_kmeans.t >>>>>>>> (Wstat: 256 Tests: 18 Failed: 1) Failed test: 13 >>>>>>>> Non-zero exit status: 1 Files=5, Tests=147, 1 wallclock >>>>>>>> secs ( 0.03 usr 0.01 sys + 0.91 >>>>>>>> >>>>>>> cusr >>>> >>>>> 0.04 csys = 0.99 CPU) >>>>>> >>>>>>> Result: FAIL Failed 1/5 test programs. 1/147 subtests >>>>>>>> failed. make: *** [test_dynamic] Error 255 -> FAIL >>>>>>>> Installing PDL::Stats failed. See >>>>>>>> >>>>>>> /Volumes/roller/Users/punkish/**.cpanm/build.log for details. >>>>>> >>>>>>> ---- >>>>>>>> >>>>>>>> >>>>>>>> Here is line 168 (test #13) from t/stats_kmeans.t >>>>>>>> >>>>>>>> 168> is(tapprox( t_kmeans_bad(), 0 ), 1); 169> sub >>>>>>>> t_kmeans_bad { 170> my $data = sequence 7, 3; 171> >>>>>>>> $data = $data->setbadat(4,0); 172> my %m = >>>>>>>> $data->kmeans({NCLUS=>2, NTRY=>10, V=>0}); 173> >>>>>>>> return sum( $m{ms}->sumover - pdl qw( 1.5 1.9166667 >>>>>>>> >>>>>>> 1.9166667 >>>> >>>>> ) ); >>>>>> >>>>>>> 174> } >>>>>>>> >>>>>>>> >>>>>>>> I am on Mac OS X Lion with Perl 5.14.1 and PDL 2.4.6 >>>>>>>> installed via >>>>>>>> >>>>>>> MacPorts >>>>>> >>>>>>> ^^^^^ ||||| >>>>>>> >>>>>>> My guess it the problem is you are using PDL-2.4.6 and not >>>>>>> the current PDL-2.4.9 (or even better the current CPAN >>>>>>> Developers release version). >>>>>>> >>>>>>> I know there were some fixes the are in PDL-2.4.9 that >>>>>>> helped Maggie X's stuff work. Maybe she can recall any >>>>>>> tricks to build her modules with a fairly old PDL (i.e., >>>>>>> one not having the many fixes and features added in the >>>>>>> past couple of years). >>>>>>> >>>>>>> --Chris >>>>>>> >>>>>>> On Dec 22, 2011, at 3:26 PM, Maggie X wrote: >>>>>>>> >>>>>>>> Hi Chris, >>>>>>>>> >>>>>>>>> Thanks for including the info! Here's the blurb for >>>>>>>>> PDL::Stats. >>>>>>>>> >>>>>>>>> -------------- This is a collection of statistics >>>>>>>>> modules in Perl Data Language, >>>>>>>>> >>>>>>>> with >>>> >>>>> a >>>>>> >>>>>>> >>>>>>>> quick-start guide for non-PDL people. >>>>>>>>> >>>>>>>>> They make perldl--the simple shell for PDL--work like a >>>>>>>>> teenie weenie >>>>>>>>> >>>>>>>> R, >>>>>> >>>>>>> >>>>>>>> but with PDL threading--"the fast (and automagic) >>>>>>>>> vectorised >>>>>>>>> >>>>>>>> iteration >>>> >>>>> of >>>>>> >>>>>>> 'elementary operations' over arbitrary slices of >>>>>>>>> multidimensional >>>>>>>>> >>>>>>>> data"--of >>>>>> >>>>>>> procedures including t-test, ordinary least squares >>>>>>>>> regression, and >>>>>>>>> >>>>>>>> k-means >>>>>> >>>>>>> clustering. --------------- >>>>>>>>> >>>>>>>>> Best, Maggie >>>>>>>>> >>>>>>>>
_______________________________________________ Perldl mailing list [email protected] http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
