Bug in GSL 0.7

Bug in GSL 0.7

[Edgar Garduño Ángeles]

I compiled the GSL under RH 7.0 and I got the following error when I ran
"make check":

PASS:   gsl_sf_eta_impl( 20, &r)
PASS: Zeta Functions
PASS: Result Methods
FAIL: test
===================
1 of 1 tests failed
===================
make[2]: *** [check-TESTS] Error 1
make[2]: Leaving directory `/home/garduno/gsl-0.7/specfunc'
make[1]: *** [check-am] Error 2
make[1]: Leaving directory `/home/garduno/gsl-0.7/specfunc'
make: *** [check-recursive] Error 1

Re: Bug in GSL 0.7

[Brian Gough]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 825 bytes --]

Thanks for the bug report. The failure has occurred somewhere further
up in the test. There should be a more detailed error message at that
point.  You may want to redirect the output into a log file to find
it -- please let us know what you find.
best regards
Brian Gough


Edgar Garduño Ángeles writes:
 > I compiled the GSL under RH 7.0 and I got the following error when I ran
 > "make check":
 > 
 > PASS:   gsl_sf_eta_impl( 20, &r)
 > PASS: Zeta Functions
 > PASS: Result Methods
 > FAIL: test
 > ===================
 > 1 of 1 tests failed
 > ===================
 > make[2]: *** [check-TESTS] Error 1
 > make[2]: Leaving directory `/home/garduno/gsl-0.7/specfunc'
 > make[1]: *** [check-am] Error 2
 > make[1]: Leaving directory `/home/garduno/gsl-0.7/specfunc'
 > make: *** [check-recursive] Error 1
 > 
 > 
 > 

Re: Bug in GSL 0.7

[Edgar Garduño Ángeles]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1152 bytes --]

I double checked my compilation and I am sorry, you are right the error
was further up. Here is what I found:

GSL 0.7 compiled on a DELL Dual Pentium III (733 MHz) with RH 7.0:

PASS: Dilogarithm
PASS:   gsl_sf_multiply_impl(-3.0,2.0, &r)
PASS:   gsl_sf_multiply_impl(x, 1.0/x, &r)
PASS:   gsl_sf_multiply_impl(x, 0.2, &r)
PASS:   gsl_sf_multiply_impl(x, 4.0, &r)
FAIL: Elementary Functions (Misc)
.
.
.
PASS: Zeta Functions
PASS: Result Methods
FAIL: test
===================
1 of 1 tests failed
===================
make[2]: *** [check-TESTS] Error 1
make[2]: Leaving directory
`/home/garduno/libraries/GNU-SL/gsl-0.7/specfunc'
make[1]: *** [check-am] Error 2
make[1]: Leaving directory
`/home/garduno/libraries/GNU-SL/gsl-0.7/specfunc'
make: *** [check-recursive] Error 1

> Thanks for the bug report. The failure has occurred somewhere further
> up in the test. There should be a more detailed error message at that
> point.  You may want to redirect the output into a log file to find
> it -- please let us know what you find.

-- 
Best regards,
________________________________________________________________________
Edgar Garduño Ángeles

Re: Bug in GSL 0.7

[Adriano Roberto de Lima]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 5780 bytes --]

Hello. I also have done "make check" (I am using SuSE 7.0) and I got the
error message

.
.
.
Testing Tsuda's function
PASS: miser(f3), dim=1, err=0.0001 (0.99995 observed vs 1 expected)
PASS: miser(f3), dim=2, err=0.0011 (1.00072 observed vs 1 expected)
PASS: miser(f3), dim=3, err=0.0031 (1.00119 observed vs 1 expected)
PASS: miser(f3), dim=4, err=0.0051 (1.00575 observed vs 1 expected)
PASS: miser(f3), dim=5, err=0.0076 (1.00606 observed vs 1 expected)
PASS: miser(f3), dim=6, err=0.0090 (0.991593 observed vs 1 expected)
PASS: miser(f3), dim=7, err=0.0122 (1.00468 observed vs 1 expected)
PASS: miser(f3), dim=8, err=0.0150 (0.983598 observed vs 1 expected)
PASS: miser(f3), dim=9, err=0.0181 (0.987409 observed vs 1 expected)
FAIL: miser_test
testing allocation/input checks
PASS: error if not initialized
PASS: error if limits too large
PASS: error if num_dim = 0
PASS: error if calls = 0
PASS: error if xu < xl
Testing constant function and normalization
PASS: plain(fconst), calls=1199, dim=1, err=0.0000 (2 observed vs 2
expected)
PASS: plain(fconst), calls=1438, dim=2, err=0.0000 (4 observed vs 4
expected)
PASS: plain(fconst), calls=1725, dim=3, err=0.0000 (8 observed vs 8
expected)
PASS: plain(fconst), calls=2069, dim=4, err=0.0000 (16 observed vs 16
expected)
PASS: plain(fconst), calls=2482, dim=5, err=0.0000 (32 observed vs 32
expected)
PASS: plain(fconst), calls=2978, dim=6, err=0.0000 (64 observed vs 64
expected)
PASS: plain(fconst), calls=3573, dim=7, err=0.0000 (128 observed vs 128
expected)
PASS: plain(fconst), calls=4287, dim=8, err=0.0000 (256 observed vs 256
expected)
PASS: plain(fconst), calls=5144, dim=9, err=0.0000 (512 observed vs 512
expected)
Testing product function
PASS: plain(f0), calls=5400, dim=1, err=0.0079 (1.01052 observed vs 1
expected)
PASS: plain(f0), calls=9720, dim=2, err=0.0090 (1.001 observed vs 1
expected)
PASS: plain(f0), calls=17496, dim=3, err=0.0088 (0.998185 observed vs 1
expected)
PASS: plain(f0), calls=31492, dim=4, err=0.0083 (0.999813 observed vs 1
expected)
PASS: plain(f0), calls=56685, dim=5, err=0.0075 (1.00412 observed vs 1
expected)PASS: plain(f0), calls=102033, dim=6, err=0.0159 (0.994534
observed vs 1 expected)
PASS: plain(f0), calls=183659, dim=7, err=0.0178 (0.991671 observed vs 1
expected)
PASS: plain(f0), calls=330586, dim=8, err=0.0393 (0.999729 observed vs 1
expected)
PASS: plain(f0), calls=595054, dim=9, err=0.0625 (1.00143 observed vs 1
expected)
Testing single gaussian
PASS: plain(f1), calls=20000, dim=1, err=0.0122 (0.993646 observed vs 1
expected)
PASS: plain(f1), calls=40000, dim=3, err=0.0384 (0.977327 observed vs 1
expected)
PASS: plain(f1), calls=120000, dim=5, err=0.3195 (1.12906 observed vs 1
expected)
PASS: plain(f1), calls=500000, dim=7, err=2.7032 (0.856804 observed vs 1
expected)
Testing double gaussian
PASS: plain(f2), calls=20000, dim=1, err=0.0071 (0.999286 observed vs 1
expected)
PASS: plain(f2), calls=40000, dim=3, err=0.0283 (1.00717 observed vs 1
expected)PASS: plain(f2), calls=100000, dim=5, err=0.1901 (1.04876
observed vs 1 expected)
PASS: plain(f2), calls=120000, dim=7, err=0.8730 (0.983047 observed vs 1
expected)
PASS: plain(f2), calls=1000000, dim=9, err=5.4075 (1.20939 observed vs 1
expected)
Testing Tsuda's function
PASS: plain(f3), calls=11999, dim=1, err=0.0064 (1.00537 observed vs 1
expected)PASS: plain(f3), calls=14398, dim=2, err=0.0094 (1.01132 observed
vs 1 expected)PASS: plain(f3), calls=17277, dim=3, err=0.0115 (0.997633
observed vs 1 expected)
PASS: plain(f3), calls=20732, dim=4, err=0.0132 (0.975394 observed vs 1
expected)
PASS: plain(f3), calls=24878, dim=5, err=0.0174 (1.00307 observed vs 1
expected)PASS: plain(f3), calls=29853, dim=6, err=0.0175 (0.973562
observed vs 1 expected)
PASS: plain(f3), calls=35823, dim=7, err=0.0228 (1.01513 observed vs 1
expected)PASS: plain(f3), calls=42987, dim=8, err=0.0220 (1.02606 observed
vs 1 expected)PASS: plain(f3), calls=51584, dim=9, err=0.0272 (1.00048
observed vs 1 expected)PASS: plain_test
===================
2 of 3 tests failed
===================
make[2]: *** [check-TESTS] Error 1
make[2]: Leaving directory `/home/arlima/gsl/gsl-0.7/monte'
make[1]: *** [check-am] Error 2
make[1]: Leaving directory `/home/arlima/gsl/gsl-0.7/monte'
make: *** [check-recursive] Error 1


/-----------------------------------+-------------------\
|     Adriano Roberto de Lima       |        .~.        |
|  PMMH/ESPCI - Paris - France      |       / v \       |
|  email: arlima at pmmh.espci.fr   |      /(   )\      |
|    email: arlima at bigfoot.com   |        ^^^        |
|           arlima at pmmh.espci.fr | I use GNU/LINUX ! |
| hpage: www.pmmh.espci.fr/~arlima  |    User #35074    |
\-------------------------------------------------------/
Lets crash some "sniffers": bomb, president, atomic, 
military, secret, espionage, drugs, carnivore animal.



On Wed, 1 Nov 2000, Brian Gough wrote:

> Thanks for the bug report. The failure has occurred somewhere further
> up in the test. There should be a more detailed error message at that
> point.  You may want to redirect the output into a log file to find
> it -- please let us know what you find.
> best regards
> Brian Gough
> 
> 
> Edgar Garduño Ángeles writes:
>  > I compiled the GSL under RH 7.0 and I got the following error when I ran
>  > "make check":
>  > 
>  > PASS:   gsl_sf_eta_impl( 20, &r)
>  > PASS: Zeta Functions
>  > PASS: Result Methods
>  > FAIL: test
>  > ===================
>  > 1 of 1 tests failed
>  > ===================
>  > make[2]: *** [check-TESTS] Error 1
>  > make[2]: Leaving directory `/home/garduno/gsl-0.7/specfunc'
>  > make[1]: *** [check-am] Error 2
>  > make[1]: Leaving directory `/home/garduno/gsl-0.7/specfunc'
>  > make: *** [check-recursive] Error 1
>  > 
>  > 
>  > 
> 

Re: Bug in GSL 0.7

[Brian Gough]

There should be a detailed error message somewhere in the output --
with the expected and observed value and relative error looking
something like this,

  FAIL: ....
     expected: 1.23456
     obtained: 1.23444  3.45e-7  2.45e-8
     fracdiff: 5.21e1
    
     value/expected not consistent within reported error

Does this appear anywhere?

Re: Bug in GSL 0.7

[Brian Gough]

Thanks -- the failures in the monte directory should be listed in the
KNOWN-PROBLEMS file.  There will be a fixed version of the monte carlo
codes in the next release.
best regards
Brian Gough