Dear Antony,
I have the following problem with CosmoMc.
If I set omech2 to the defaults value it's all ok.
But when I try to fix a given value like omech2 = 0.15, CosmoMC crashes because the Cholesky Matrix is not positive definite.
Would you help me?
Thanks a lot in advance!
Fast divided into 1 blocks
19 parameters ( 5 slow ( 2 semi-slow), 14 fast ( 0 semi-fast))
skipped unused params: bm_1_2 bm_1_3 bm_1_4 bm_1_5 bm_2_1 bm_2_2 bm_2_3 bm_2_4 bm_2_5 bm_3_1 bm_3_2 bm_3_3 bm_3_4 bm_3_5 bm_4_1 bm_4_2 bm_4_3 bm_4_4 bm_4_5
1 Reading checkpoint from chains/test1_1.chk
Matrix_Cholesky: not positive definite 3
MpiStop: 0
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
with errorcode 0.
NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
Cholesky Matrix Error with CosmoMc
-
- Posts: 15
- Joined: April 01 2015
- Affiliation: University of Rome - Tor Vergata
-
- Posts: 1941
- Joined: September 23 2004
- Affiliation: University of Sussex
- Contact:
Re: Cholesky Matrix Error with CosmoMc
Delete your checkpoint files *.chk and try again?
-
- Posts: 15
- Joined: April 01 2015
- Affiliation: University of Rome - Tor Vergata
Cholesky Matrix Error with CosmoMc
Dear Antony it seems that is working and MonteCarlo process has begun without problems!
Thank you again for your help Antony,
Best regards
Thank you again for your help Antony,
Best regards