[Met_help] FW: Stat_Analysis
John Halley Gotway
johnhg at rap.ucar.edu
Thu Mar 11 13:14:36 MST 2010
Bob,
Since I'm at home with my son today, I'm working on a PC and don't have
access to running MET on your data. I can give you some suggestions to
try, and if it's still an issue tomorrow, I'll be in the office and will
be able to diagnose the problem better.
The Stat-Analysis tool doesn't do anything fancy. It does exact string
matching with the data in the STAT files. I'm guessing one of your
filtering options doesn't EXACTLY match what's in the STAT files. Let's
try running the most basic job, and then add in the filtering options one
at a time to diagnose the problem.
Please try running the following job on the command line with no config file:
/sphome/qcteam/METv2.0/bin/stat_analysis \
-lookin /sphome/qcteam/METv2.0/out/grid_stat/LIS \
-job aggregate_stat -line_type FHO -out_line_type CTS \
-dump_row -dump_row ../out/stat_analysis/job_aggregate_CTC.stat
That will probably match a bunch of lines. Next, rerun that same job many
times, each time adding one more of the following filtering options:
-model LIS_Q-MORPH
-fcst_lead 030000
-fcst_valid_beg 20100301_030000
-fcst_valid_end 20100308_030000
-fcst_init_beg 20100301_030000
-fcst_init_end 20100308_030000
-fcst_init_hour 000000
-fcst_var APCP/3
-fcst_lev A03
-obs_lev A03
-obtype MC_PCP
-vx_mask FULL
Just keep adding those one at a time until the number of matching lines
goes down to zero. That will tell you which filtering option is the
culprit. Then adjust it as necessary to exactly match what's in the STAT
output files.
I usually run these types of STAT-Analysis jobs on the command line until
I get them doing what I want, and then I put them into a config file. Let
me know if you're still having trouble tomorrow, and I can take a closer
look.
Thanks,
John
> aggregate_stat -model LIS_Q-MORPH -fcst_lead 030000 -fcst_val
> id_beg 20100301_030000 -fcst_valid_end 20100308_030000 -fcst_init_beg
> 20100301_030000 -fcst_init_end 20100308_030000 -fcst_init_hour 00
> 0000 -fcst_var APCP/3 -fcst_lev A03 -obs_lev A03 -obtype MC_PCP -vx_mask
> FULL -alpha 0.000000 -line_type FHO -dump_row ../out/stat_anal
> ysis/job_aggregate_SL1L2.stat -out_line_type CTS -out_alpha 0.050000
> John, I changed "LIS Q-MORPH" to "LIS_Q-MORPH" and reran the grid stat
> verification. The following is the output from one of the ctc files:
>
> VERSION MODEL FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD
> OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_LEV OBS_VAR OBS_LEV
> OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH
> ALPHA LINE_TYPE TOTAL FY_OY FY_ON FN_OY FN_ON
> V2.0 LIS_Q-MORPH 030000 20100308_030000 20100308_030000 000000
> 20100308_030000 20100308_030000 APCP_03 A03 APCP_03 A03
> MC_PCP FULL UW_MEAN 1 >0.000 >0.000 NA
> NA
> CTC 171679 303 9216 1131 161029
> V2.0 LIS_Q-MORPH 030000 20100308_030000 20100308_030000 000000
> 20100308_030000 20100308_030000 APCP_03 A03 APCP_03 A03
> MC_PCP FULL UW_MEAN 1 >=5.000 >=5.000 NA
> NA
> CTC 171679 0 0 4 171675
>
> My config file settings are attached.
>
> Below is the output returned:
>
> WARNING: do_job_aggr_stat() -> no matching STAT lines found for job: -job
> aggregate_stat -model LIS_Q-MORPH -fcst_lead 030000 -fcst_val
> id_beg 20100301_030000 -fcst_valid_end 20100308_030000 -fcst_init_beg
> 20100301_030000 -fcst_init_end 20100308_030000 -fcst_init_hour 00
> 0000 -fcst_var APCP/3 -fcst_lev A03 -obs_lev A03 -obtype MC_PCP -vx_mask
> FULL -alpha 0.000000 -line_type FHO -dump_row ../out/stat_anal
> ysis/job_aggregate_SL1L2.stat -out_line_type CTS -out_alpha 0.050000
> Creating STAT output file "../out/stat_analysis/job_aggregate_SL1L2.stat"
> WARNING: do_job_aggr_stat() -> no matching STAT lines found for job: -job
> aggregate_stat -model LIS_Q-MORPH -fcst_lead 030000 -fcst_val
> id_beg 20100301_030000 -fcst_valid_end 20100308_030000 -fcst_init_beg
> 20100301_030000 -fcst_init_end 20100308_030000 -fcst_init_hour 00
> 0000 -fcst_var APCP/3 -fcst_lev A03 -obs_lev A03 -obtype MC_PCP -vx_mask
> FULL -alpha 0.000000 -line_type CTC -dump_row ../out/stat_anal
> ysis/job_aggregate_SL1L2.stat -out_line_type CTS -out_alpha 0.050000
>
> So, why can't it find a match? What am I missing?
>
> Thanks
>
> Bob
>
>
> -----Original Message-----
> From: John Halley Gotway [mailto:johnhg at rap.ucar.edu]
> Sent: Tuesday, March 09, 2010 3:11 PM
> To: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Cc: met_help at ucar.edu
> Subject: Re: FW: Stat_Analysis
>
> Bob,
>
> I took a look at the files you sent, and I see what's going on. It looks
> like it was just a poor choice of MODEL name when you ran Grid-Stat. You
> chose the model name "LIS Q-MORPH" with an embedded space.
>
> It ran through Grid-Stat fine, but when STAT-Analysis parses that STAT
> data it gets confused by the embedded space. I tried changing that model
> name to "LISQ-MORPH" in the data you sent, and rerunning a STAT-Analysis
> job, and it worked fine.
>
> I'm surprised Grid-Stat allowed you to use a MODEL name with an embedded
> space. I thought it was checking for embedded whitespace and erroring out
> if it found any.
>
> I'll take a closer look at that.
>
> Hope that solves the problem.
>
> Thanks,
> John
>
>>
>>
>> John, until I can get the GUI figured out, I have been trying to do this
>> via
>> command line, and not having much luck. The configuration file is
>> attached.
>>
>>
>>
>> The command line to kick it off is:
>> /sphome/qcteam/METv2.0/bin/stat_analysis
>> -lookin /sphome/qcteam/METv2.0/out/grid_stat/LIS -config
>> /sphome/qcteam/METv2.0/scripts/config/STATAnalysisConfig_LIS
>>
>>
>>
>> Also attached are the data files. I didn't include the nc files since
>> they
>> were 11 mb each. I ran verification of APCP for a four day period. I
>> had
>> to rename the zip file to met_data.dat to get it past our firewall.
>>
>>
>>
>> Below is the typical error message a get:
>>
>>
>>
>> Creating STAT output file
>> "../out/stat_analysis/job_aggregate_SL1L2.stat"
>>
>> WARNING: do_job_aggr_stat() -> no matching STAT lines found for job:
>> -job
>> aggregate_stat -model LIS -fcst_lead 030000 -fcst_valid_beg 20
>>
>> 100301_000000 -fcst_valid_end 20100304_000000 -fcst_init_hour 000000
>> -fcst_var APCP/3 -fcst_lev SFC -obs_lev SFC -alpha 0.000000 -line_t
>>
>> ype FHO -dump_row ../out/stat_analysis/job_aggregate_SL1L2.stat
>> -out_line_type CTS -out_alpha 0.050000
>>
>>
>>
>> I have tried a lot of different options and combinations but I can't get
>> past this error. The verification files are in the proper directory
>> (lookin
>> above)
>>
>>
>>
>> Again, what have I done wrong?
>>
>>
>>
>> Thanks
>>
>>
>>
>> Bob
>>
>>
>
>
>
More information about the Met_help
mailing list