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
