bbp
bbp copied to clipboard
SCECcode
GP HF method fails intermittently
I'm experiencing intermittent failures on the HF part of the GP calculation with the attached inputs. I originally believed that it was dependent on the seed, but failure is intermittent with the same seed. The attached inputs failed in 2 of 10 tries.
Current dev branch on Xubuntu 16.4.
Here is the output:
------------------------------------GP Jbsim------------------------------------
==> Generating LF seismogram for station: USC
==> Generating LF seismogram for station: SBSM
-------------------------------GP Jbsim Completed-------------------------------
-----------------------------------GP HfSims------------------------------------
Running: /home/kevin/bbp/dev/bbp/src/gp/bin/srf2stoch infile=/home/kevin/bbp/bbp_data/indata/6121001/event_145982_0.1s_CONST_VEL_ADJ_LEN.srf outfile=/home/kevin/bbp/bbp_data/tmpdata/6121001/event_145982_0.1s_CONST_VEL_ADJ_LEN.srf.stations_cs_sites.2.000000x2.000000 target_dx=2.000000 target_dy=2.000000 >> /home/kevin/bbp/bbp_data/logs/6121001/6121001.hfsims_stations_cs_sites.log 2>&1
Running: /home/kevin/bbp/dev/bbp/src/gp/bin/hb_high_v5.4.3 >> /home/kevin/bbp/bbp_data/logs/6121001/6121001.hfsims_stations_cs_sites.log 2>&1 << END
50
/home/kevin/bbp/bbp_data/indata/6121001/stations_cs_sites.stl
/home/kevin/bbp/bbp_data/tmpdata/6121001/tmp_hfsim_out
2 1 2
1
4 0 0.02 19.9
12345523
2
102.400000 0.010000 10.000000 0.040000 0.600000
0.841246 0.700000 1.000000 0.000000 0.100000
-1 -1.000000
/home/kevin/bbp/bbp_data/tmpdata/6121001/event_145982_0.1s_CONST_VEL_ADJ_LEN.srf.stations_cs_sites.2.000000x2.000000
/home/kevin/bbp/bbp_data/indata/6121001/genslip_nr_generic1d-gp01.vmod_stations_cs_sites.local
999.900000
-99 0.0 0.0 0.0 0.0 1
-1
0.000000 0.0 0.000000
0
END
------------------------------GP HfSims Completed-------------------------------
-------------------------------------Match--------------------------------------
==> Processing velocity seismograms for station: USC
==> Processing acceleration seismograms for station: USC
Traceback (most recent call last):
File "/home/kevin/bbp/dev/bbp/comps/run_bbp.py", line 295, in <module>
obj.run()
File "/home/kevin/bbp/dev/bbp/comps/match.py", line 465, in run
print_cmd=False)
File "/home/kevin/bbp/dev/bbp/comps/bband_utils.py", line 85, in runprog
(cmd, proc.returncode))
BroadbandExternalError: None
/home/kevin/bbp/dev/bbp/src/gp/bin/wcc_resamp_arbdt newdt=0.010000 pow2=0 infile=/home/kevin/bbp/bbp_data/tmpdata/6121001/6121001.USC-hf.acc.000 outfile=/home/kevin/bbp/bbp_data/tmpdata/6121001/6121001.USC-hf-resamp.000 >> /home/kevin/bbp/bbp_data/logs/6121001/6121001.match_stations_cs_sites.log 2>&1 returned 139
My original SRF files were a little wonky, turns out that formatting matters. Fixed the SRFs and now have realistic ground motions when it works, but still getting the same intermittent failures. Here is a new set of input files and logs that sometimes fail: bbp_hf_error.zip