Dear Alfonso,
I did the two things you suggested and tested several scenarios, please see below.
Could you let me know what I should do next to get where the bug is coming from?
Thanks a lot,
Sophie

-------------------------

Everything was run with config -t 10:00:00 -m 8Gb


####I changed the preferences as suggested on the png#####
1) Running my full .m supposed to create the .mat file and run it in conn, with nsubjects=1015/ncore=100;

matlab -nodesktop -nojvm -batch 'run LesionNetworkMappingSoso_SCITAS_17122020v5.m'
after waiting 40min, nothing was happening, no jobs were submitted. (I imagine that if I had waited more, I would have had the bus error below)


2) Creating the .mat manually while debugging and then, run conn_batch(BATCH), with nsubjects=1015/ncore=100;

matlab -nodesktop -nojvm -batch 'run LesionNetworkMappingSoso_SCITAS_17122020v6.m'
parallelization settings loaded from /home/betka/Script_SCITAS/conn/conn_jobmanager.mat
and, after 40min, nothing is submitted. (I imagine that if I had waited more, I would have had the bus error below)





####Then I changed the conn.m (while keeping the preferences as on the png)####
1) Running my full .m supposed to create the .mat file and run it in conn, with nsubjects=1015/ncore=100;

matlab -nodesktop -nojvm -batch 'run LesionNetworkMappingSoso_SCITAS_17122020v5.m'
Warning: Directory already exists.
> In LesionNetworkMappingSoso_SCITAS_17122020v5 (line 11)
In run (line 91)

and after sometimes:
Bus error (core dumped)


2) Creating the .mat manually while debugging and then, run conn_batch(BATCH), with nsubjects=1015/ncore=100;

matlab -nodesktop -nojvm -batch 'run LesionNetworkMappingSoso_SCITAS_17122020v6.m'
parallelization settings loaded from /home/betka/Script_SCITAS/conn/conn_jobmanager.mat

then after a long time:
Bus error (core dumped)


3) As a sanity check I tried to use the same .m as 1) but with nsubjects=10/ncore=100; this works perfectly.