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?
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
