Source: octave-sparsersb Version: 1.0.6-1 User: debian...@lists.debian.org Usertags: needs-update
Dear Octave maintainers, Recently octave-sparsersb started to fail (but some versions of 1.0.6-1 passed) its autopgktest on ci.debian.net¹. I copied the error below. Unfortunately, whatever caused the regression, it seems it already migrated to testing, as references also started to fail there. Could you please investigate and either fix the autopkgtest or reassign to the offending package? Paul ¹ https://ci.debian.net/packages/o/octave-sparsersb/unstable/amd64/ https://ci.debian.net/data/autopkgtest/unstable/amd64/o/octave-sparsersb/568398/log.gz autopkgtest [11:51:26]: test command1: [----------------------- Checking package... Checking m files ... Checking C++ files ... [sparsersb.cc] >>>>> /tmp/autopkgtest-lxc.byt8xejv/downtmp/build.uz2/src/src/sparsersb.cc ***** test #help sparsersb s=sparsersb([2]), assert(s==2), assert(s!=1) !!!!! test failed 'sparsersb' undefined near line 4 column 4 ***** test s=sparsersb([1,2],[1,1],[11,21],2,2 ), assert(nnz(s)==2) !!!!! test failed 'sparsersb' undefined near line 3 column 4 ***** test s=sparsersb([1,2],[1,1],[11,21],2,2,-1 ), assert(nnz(s)==2) !!!!! test failed 'sparsersb' undefined near line 3 column 4 ***** test s=sparsersb([1,2],[1,1],[11,21] ), assert(nnz(s)==2) !!!!! test failed 'sparsersb' undefined near line 3 column 4 ***** test s=sparsersb(10,10 ), assert(nnz(s)==0) !!!!! test failed 'sparsersb' undefined near line 3 column 4 ***** test s=sparsersb([1,1],[1,1],[11,21] ), assert(nnz(s)==1), assert(s(1,1)==32) !!!!! test failed 'sparsersb' undefined near line 3 column 4 ***** test s=sparsersb([1,1],[1,1],[11,21],2,2,"unique"), assert(nnz(s)==1), assert(s(1,1)==21) !!!!! test failed 'sparsersb' undefined near line 3 column 4 ***** test sparsersb("set","RSB_IO_WANT_VERBOSE_TUNING","1") !!!!! test failed 'sparsersb' undefined near line 3 column 2 ***** test # sparsersb("get","RSB_IO_WANT_VERBOSE_TUNING","1") # FIXME ***** test sparsersb(sparsersb([11,0;21,22]),"RSB_MIF_TOTAL_SIZE__TO__SIZE_T") !!!!! test failed 'sparsersb' undefined near line 3 column 12 ***** test sparsersb(sparsersb([11,0;21,22]),"save","sparsersb_temporary_matrix_file.mtx") !!!!! test failed 'sparsersb' undefined near line 3 column 12 ***** test [S, NROWS, NCOLS, NNZ, REPINFO, FIELD, SYMMETRY] = sparsersb("sparsersb_temporary_matrix_file.mtx" ); assert(NROWS==2);assert(NCOLS==2);assert(NNZ==3);assert(FIELD=="real" );assert(SYMMETRY=='U'); !!!!! test failed 'sparsersb' undefined near line 3 column 46 ***** test [S, NROWS, NCOLS, NNZ, REPINFO, FIELD, SYMMETRY] = sparsersb("sparsersb_temporary_matrix_file.mtx", "Z"); assert(NROWS==2);assert(NCOLS==2);assert(NNZ==3);assert(FIELD=="complex");assert(SYMMETRY=='U'); !!!!! test failed 'sparsersb' undefined near line 3 column 46 ***** test rrm=sparsersb(sprand(1000,1000,0.001)); sparsersb(rrm,"render", "sparsersb_temporary_render.eps" ,1024,1024); # sparsersb(rrm,"renderb", "sparsersb_temporary_renderb.eps"); sparsersb(rrm,"renders", "sparsersb_temporary_renders.eps"); # FIXME !!!!! test failed 'sparsersb' undefined near line 3 column 6 ***** test sparsersb(sparsersb(sprand(100,100,0.4)),"autotune","n",20,4,1,1,1) !!!!! test failed 'sparsersb' undefined near line 3 column 12 ***** demo # You can use 'sparsersb' just like 'sparse' in the most of cases: R=(rand(3)>.6) # R = # # 0 0 0 # 0 0 0 # 1 0 1 # A_octave=sparse(R) # A_octave = # # Compressed Column Sparse (rows = 3, cols = 3, nnz = 2 [22%]) # # (3, 1) -> 1 # (3, 3) -> 1 # A_librsb=sparsersb(R) # A_librsb = # # Recursive Sparse Blocks (rows = 3, cols = 3, nnz = 2, symm = U [22%]) # # (3, 1) -> 1 # (3, 3) -> 1 # # test sparsersb # ... # help sparsersb ***** demo # The interface of 'sparsersb' is almost like the one of 'sparse'. sparsersb([2]); # 1x1 matrix sparsersb([1,2],[1,1],[11,21] ); # 2x1 matrix sparsersb([1,2],[1,1],[11,21],2,2); # 2x2 matrix sparsersb([1,2,2 ],[1,1,2 ],[11,21, 22],2,2); # 2x2 lower triangular sparsersb([1,2,2,2],[1,1,2,2],[11,21,11,11],2,2); # 2x2 lower triangular, last element ignored sparsersb([1,2,2,2],[1,1,2,2],[11,21,11,11],2,2,"unique"); # 2x2 lower triangular, last element ignored sparsersb([1,2,2,2],[1,1,2,2],[11,21,11,11],2,2,"sum"); # 2x2 lower triangular, last two elements summed # But it has a extensions, like symmetric and hermitian matrices. sparsersb([1,2,2 ],[1,1,2 ],[11,21 , 22],2,2,"general"); # 2x2 lower tringular sparsersb([1,2,2 ],[1,1,2 ],[11,21 , 22],2,2,"symmetric"); # 2x2 symmetric (only lower triangle stored) sparsersb([1,2,2 ],[1,1,2 ],[11,21i, 22],2,2,"hermitian"); # 2x2 hermitian (only lower triangle stored) ***** demo # Any 'sparse' or 'dense' matrix can be converted to 'sparsersb'. d=sparsersb( [1,2;3,4] ); s=sparsersb(sparse([1,2;3,4])); # Many matrix operators are active, e.g.: +,*,-,/,\ among others... s+d; s*d; s-d; s/d; s\[1;1]; # ... ***** demo # On large matrices 'sparsersb' is supposed to be faster than 'sparse' in sparse matrix-vector multiplication: M=10000;N=10000;P=100 / M; s=sparse(sprand(M,N,P)); r=sparsersb(s); x=ones(M,1); assert(nnz(s)==nnz(r)) printf("Here, a %.2e x %.2e matrix with %.2e nonzeroes.\n",M,N,nnz(s)) tic(); sc=0; while(toc()<3) s*x; sc=sc+1; endwhile st=toc()/sc; printf("Each multiplication with 'sparse' took %.1es.\n",st); tic(); rc=0; while(toc()<3) r*x; rc=rc+1; endwhile rt=toc()/rc; printf("Each multiplication with 'sparsersb' took %.3es, this is %.4g%% of the time taken by 'sparse'.\n",rt,100*rt/st); # 'sparsersb' has an 'empirical online auto-tuning' function nsb=str2num(sparsersb(r,"get","RSB_MIF_LEAVES_COUNT__TO__RSB_BLK_INDEX_T")); # after 'autotuning' for a specific operation, this will perform faster tic; r=sparsersb(r,"autotune","n",1); toc; nnb=str2num(sparsersb(r,"get","RSB_MIF_LEAVES_COUNT__TO__RSB_BLK_INDEX_T")); printf ("Autotuning took %.2es (%d -> %d RSB blocks).\n", toc, nsb, nnb); tic(); rc=0; while(toc()<3) r*x; rc=rc+1; endwhile rt=toc()/rc; printf("Each 'optimized' multiplication with 'sparsersb' took %.3es, this is %.4g%% of the time taken by 'sparse'.\n",rt,100*rt/st); ***** demo # 'sparsersb' can render matrices into Encapsulated Postscript files: rm = sparsersb(sprand(100000,100000,.0001)); sparsersb(rm,'render','sptest.eps') 15 tests, 1 passed, 0 known failure, 0 skipped Run tests in debian/check.m make: Entering directory '/tmp/autopkgtest-lxc.byt8xejv/downtmp/build.uz2/src/src' make: -p: Command not found LFLAGS=" " CXXFLAGS=" " -D'RSB_SPARSERSB_LABEL=sparsersb' -o sparsersb.oct sparsersb.cc /bin/sh: 1: -DRSB_SPARSERSB_LABEL=sparsersb: not found make: *** [Makefile:22: sparsersb.oct] Error 127 make: Leaving directory '/tmp/autopkgtest-lxc.byt8xejv/downtmp/build.uz2/src/src' Summary: 16 tests, 2 passed, 0 known failures, 0 skipped Some tests failed. Giving up... autopkgtest [11:51:27]: test command1: -----------------------]
signature.asc
Description: OpenPGP digital signature
