You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I run the example through valgrind, and it caught a use of unitialised variables in MatMult from line 490 in ALStructure::run():
andrea@laptopX1:~/git/scope/examples$ valgrind --leak-check=full --track-origins=yes ../debug/scope -g ./source_files/example_1k -seed 12345 -m 10
==91512== Memcheck, a memory error detector
==91512== Copyright (C) 2002-2022, and GNU GPL'd, by Julian Seward et al.
==91512== Using Valgrind-3.22.0 and LibVEX; rerun with -h for copyright info
==91512== Command: ../debug/scope -g ./source_files/example_1k -seed 12345 -m 10
==91512==
done reading
Running on Dataset of 1000 SNPs and 1000 Individuals
==91512== Conditional jump or move depends on uninitialised value(s)
==91512== at 0x48CD8F5: __ieee754_pow_fma (e_pow.c:292)
==91512== by 0x488EE47: pow@@GLIBC_2.29 (w_pow_template.c:32)
==91512== by 0x147B57: MatMult::MatMult(Genotype&, Eigen::Matrix<double, -1, -1, 1, -1, -1>&, bool, bool, bool, bool, bool, int, int) (in /home/andrea/git/scope/debug/scope)
==91512== by 0x128DCC: ALStructure::run() (in /home/andrea/git/scope/debug/scope)
==91512== by 0x114E18: main (in /home/andrea/git/scope/debug/scope)
==91512== Uninitialised value was created by a stack allocation
==91512== at 0x114DEC: main (in /home/andrea/git/scope/debug/scope)
==91512==
==91512== Conditional jump or move depends on uninitialised value(s)
==91512== at 0x48CDAED: __ieee754_pow_fma (e_pow.c:299)
==91512== by 0x488EE47: pow@@GLIBC_2.29 (w_pow_template.c:32)
==91512== by 0x147B57: MatMult::MatMult(Genotype&, Eigen::Matrix<double, -1, -1, 1, -1, -1>&, bool, bool, bool, bool, bool, int, int) (in /home/andrea/git/scope/debug/scope)
==91512== by 0x128DCC: ALStructure::run() (in /home/andrea/git/scope/debug/scope)
==91512== by 0x114E18: main (in /home/andrea/git/scope/debug/scope)
==91512== Uninitialised value was created by a stack allocation
==91512== at 0x114DEC: main (in /home/andrea/git/scope/debug/scope)
==91512==
==91512== Conditional jump or move depends on uninitialised value(s)
==91512== at 0x48CDD63: __ieee754_pow_fma (e_pow.c:301)
==91512== by 0x488EE47: pow@@GLIBC_2.29 (w_pow_template.c:32)
==91512== by 0x147B57: MatMult::MatMult(Genotype&, Eigen::Matrix<double, -1, -1, 1, -1, -1>&, bool, bool, bool, bool, bool, int, int) (in /home/andrea/git/scope/debug/scope)
==91512== by 0x128DCC: ALStructure::run() (in /home/andrea/git/scope/debug/scope)
==91512== by 0x114E18: main (in /home/andrea/git/scope/debug/scope)
==91512== Uninitialised value was created by a stack allocation
==91512== at 0x114DEC: main (in /home/andrea/git/scope/debug/scope)
==91512==
Performing latent subspace estimation
Latent subspace esimation completed after 19 iterations
Initializing Phat using seed 12345
Iteration 2 -- RMSE 0.104633192581442
I had a quick go at feeding scope with a text input of the example dataset, to see whether the problem was related with the conversion from .bed. The text input did not raise any issue in MatMult, so I suspect something goes wrong when parsing the binary genotype codes (as a note, to get the txt option to work, I had to comment out the line computing pj, as it didn't like the text line, but that is not an issue with the example as there are no missing values; as the txt option is not exposed to the user, I presume that part of the code is deprecated in any case, I just used it as a quick test).
The text was updated successfully, but these errors were encountered:
I run the example through valgrind, and it caught a use of unitialised variables in MatMult from line 490 in ALStructure::run():
I had a quick go at feeding
scope
with a text input of the example dataset, to see whether the problem was related with the conversion from .bed. The text input did not raise any issue in MatMult, so I suspect something goes wrong when parsing the binary genotype codes (as a note, to get the txt option to work, I had to comment out the line computingpj
, as it didn't like the text line, but that is not an issue with the example as there are no missing values; as the txt option is not exposed to the user, I presume that part of the code is deprecated in any case, I just used it as a quick test).The text was updated successfully, but these errors were encountered: