[Met_help] [rt.rap.ucar.edu #82076] History for ERROR : AsciiHeader::read() -> trouble reading file
John Halley Gotway via RT
met_help at ucar.edu
Tue Jul 9 11:38:40 MDT 2019
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hello,
I'm trying to use MET-TC to do some rapid intensification verification.
Here is the command I'm issuing:
$> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
which returns:
DEBUG 1: [Source 1] Lookin: test.dat
DEBUG 1: Config File Default:
/usr/local/share/met/config/TCStatConfig_default
DEBUG 1: Config File User: TCStat_DTOPS_Config
DEBUG 2:
DEBUG 2: Processing Job 1: -job rirw -amodel DTOP -amodel DTOP -bmodel BEST
-storm_id AL152017 -basin AL -cyclone 15 -rirw_time 240000 -rirw_exact true
-rirw_thresh >=30.0 -rirw_window 000000 000000 -out_line_type CTC
-out_line_type CTS -out_alpha 0.05000
ERROR :
ERROR : AsciiHeader::read() -> trouble reading file:
ERROR : /usr/local/share/met/table_files/met_header_columns_AL,.txt
ERROR :
My config file (TCStat_DTOPS_Config) and my data file (test.dat) are
attached.
It seems MET is not properly identifying the header file to use. Could I
get some help troubleshooting this issue?
To clarify, on my system, the path /usr/local/share/met/table_files/
contains these met_header_columns_VX.X.txt files:
met_header_columns_V2.0.txt met_header_columns_V3.0.txt
met_header_columns_V4.0.txt met_header_columns_V4.2.txt
met_header_columns_V5.1.txt met_header_columns_V6.0.txt
met_header_columns_V3.0.1.txt met_header_columns_V3.1.txt
met_header_columns_V4.1.txt met_header_columns_V5.0.txt
met_header_columns_V5.2.txt
Thanks,
Matt
*Matthew Onderlinde*Meteorologist/Programmer
NHC/Technology & Science Branch
(o) 305.229.4466
<http://innovim.com/wp-content/uploads/BD-F-002a-Email-Signature-Block-for-Gmail.html#>
Matthew.Onderlinde at noaa.gov
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: ERROR : AsciiHeader::read() -> trouble reading file
From: John Halley Gotway
Time: Thu Sep 21 10:53:19 2017
Matt,
Thanks for sending some sample data. That makes it a lot easier to
debug.
The test.dat file you sent contains track data in ATCF format. While
the
tc_pairs tools is designed to read that data, the tc_stat tool is not.
tc_stat reads the ASCII output of tc_pairs.
Looking at the test.dat file, I'm guessing that you're trying to
compare
the DTOP track to the BEST track, looking for instances of rapid
intensification. If that's not the case, please let me know.
But looking at the DTOP track data, I'm concerned about the 4th column
with
RI in it, and columns 9 and 10 don't have wind speeds and pressure
values
in the range I'd expect. My guess is that these RI lines contain 24-
hour
differences, but the logic in MET hasn't been set up to process this
type
of data.
Instead, MET expects to process raw ATCF track data. Let me work
through
an example...
(1) Use the BEST track for Maria that you sent.
grep BEST test.dat > bal152017.dat
(2) Pull the ADECK tracks for Maria:
wget http://ftp.nhc.noaa.gov/atcf/aid_public/aal152017.dat.gz
gunzip aal152017.dat.gz
(3) Run the MET tc_pairs tool to match up the forecast/best tracks:
met-6.0/bin/tc_pairs -adeck aal152017.dat -bdeck bal152017.dat
-config
TCPairsConfig -out al152017_tcpairs
Running using the default TCPairs configuration file results in 2,306
forecast tracks being processed.
(4) Run the MET tc_stat tool to look for RI events but only for the
OFCL
forecast:
met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw -amodel
OFCL
-out_line_type CTC
Here's the 2x2 contingency table counts:
TOTAL FY_OY FY_ON FN_OY FN_ON: 60 1 0 21 38
Out of 60 points, OFCL got the RI correct once, it missed it 21 times,
and
38 times no RI was forecast or occurring.
(5) Now rerun the same job for ALL models present using the "-by
AMODEL"
option:
met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw -by AMODEL
-out_line_type CTC -out al152017_rirw_ctc.out
And write the results to an output file. You'll see that we now have
RI
contingency table counts computed separately for each mode.
There's many more options here you could play with:
- Set -out_line_type CTC,CTS,MPR to dump out the contingency table
counts
(CTC), statistics (CTS), and very detailed information for each time
point.
- Specify more columns for stratifying the data... add "-by LEAD" to
break
results down by lead time.
- Change the definition of RI that you'd like to apply. Perhaps, OFCL
would score better when defining RI as a 25kt increase instead of 30?
Hope this example helps get you going using the MET-TC tools to
quantify RI.
Please let us know what other questions you have.
Thanks,
John Halley Gotway
On Thu, Sep 21, 2017 at 8:50 AM, Matthew Onderlinde - NOAA Federal via
RT <
met_help at ucar.edu> wrote:
>
> Thu Sep 21 08:50:25 2017: Request 82076 was acted upon.
> Transaction: Ticket created by matthew.onderlinde at noaa.gov
> Queue: met_help
> Subject: ERROR : AsciiHeader::read() -> trouble reading file
> Owner: Nobody
> Requestors: matthew.onderlinde at noaa.gov
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
>
>
> Hello,
>
> I'm trying to use MET-TC to do some rapid intensification
verification.
> Here is the command I'm issuing:
>
> $> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
>
> which returns:
>
> DEBUG 1: [Source 1] Lookin: test.dat
> DEBUG 1: Config File Default:
> /usr/local/share/met/config/TCStatConfig_default
> DEBUG 1: Config File User: TCStat_DTOPS_Config
> DEBUG 2:
> DEBUG 2: Processing Job 1: -job rirw -amodel DTOP -amodel DTOP
-bmodel BEST
> -storm_id AL152017 -basin AL -cyclone 15 -rirw_time 240000
-rirw_exact true
> -rirw_thresh >=30.0 -rirw_window 000000 000000 -out_line_type CTC
> -out_line_type CTS -out_alpha 0.05000
> ERROR :
> ERROR : AsciiHeader::read() -> trouble reading file:
> ERROR : /usr/local/share/met/table_files/met_header_columns_AL,.txt
> ERROR :
>
> My config file (TCStat_DTOPS_Config) and my data file (test.dat) are
> attached.
>
> It seems MET is not properly identifying the header file to use.
Could I
> get some help troubleshooting this issue?
>
> To clarify, on my system, the path /usr/local/share/met/table_files/
> contains these met_header_columns_VX.X.txt files:
>
> met_header_columns_V2.0.txt met_header_columns_V3.0.txt
> met_header_columns_V4.0.txt met_header_columns_V4.2.txt
> met_header_columns_V5.1.txt met_header_columns_V6.0.txt
> met_header_columns_V3.0.1.txt met_header_columns_V3.1.txt
> met_header_columns_V4.1.txt met_header_columns_V5.0.txt
> met_header_columns_V5.2.txt
>
> Thanks,
>
> Matt
>
>
> *Matthew Onderlinde*Meteorologist/Programmer
> NHC/Technology & Science Branch
> (o) 305.229.4466 <(305)%20229-4466>
> <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> Signature-Block-for-Gmail.html#>
> Matthew.Onderlinde at noaa.gov
>
>
------------------------------------------------
Subject: ERROR : AsciiHeader::read() -> trouble reading file
From: Matthew Onderlinde - NOAA Federal
Time: Thu Sep 21 11:08:00 2017
Hi John,
Thanks so much for taking the time to help me out. I appreciate your
detailed response. I've been asked to do verification of RI
probability
forecasts. So my test.dat file contained an e-deck formatted
probability
forecast and the best track data. What probably wasn't acceptable to
MET
was catting the e-deck and the best track into one file. Regardless
of
that, I get the impression that MET isn't set up to do quantitative
verification of RI probability forecasts. Can MET produce brier
scores for
example? If not, I may just need to write the code myself.
Thanks again for your help,
Matt
*Matthew Onderlinde*Meteorologist/Programmer
NHC/Technology & Science Branch
(o) 305.229.4466
<http://innovim.com/wp-content/uploads/BD-F-002a-Email-Signature-
Block-for-Gmail.html#>
Matthew.Onderlinde at noaa.gov
On Thu, Sep 21, 2017 at 4:53 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Matt,
>
> Thanks for sending some sample data. That makes it a lot easier to
debug.
>
> The test.dat file you sent contains track data in ATCF format.
While the
> tc_pairs tools is designed to read that data, the tc_stat tool is
not.
> tc_stat reads the ASCII output of tc_pairs.
>
> Looking at the test.dat file, I'm guessing that you're trying to
compare
> the DTOP track to the BEST track, looking for instances of rapid
> intensification. If that's not the case, please let me know.
>
> But looking at the DTOP track data, I'm concerned about the 4th
column with
> RI in it, and columns 9 and 10 don't have wind speeds and pressure
values
> in the range I'd expect. My guess is that these RI lines contain
24-hour
> differences, but the logic in MET hasn't been set up to process this
type
> of data.
>
> Instead, MET expects to process raw ATCF track data. Let me work
through
> an example...
>
> (1) Use the BEST track for Maria that you sent.
>
> grep BEST test.dat > bal152017.dat
>
> (2) Pull the ADECK tracks for Maria:
>
> wget http://ftp.nhc.noaa.gov/atcf/aid_public/aal152017.dat.gz
>
> gunzip aal152017.dat.gz
> (3) Run the MET tc_pairs tool to match up the forecast/best tracks:
>
> met-6.0/bin/tc_pairs -adeck aal152017.dat -bdeck bal152017.dat
-config
> TCPairsConfig -out al152017_tcpairs
>
> Running using the default TCPairs configuration file results in
2,306
> forecast tracks being processed.
>
> (4) Run the MET tc_stat tool to look for RI events but only for the
OFCL
> forecast:
>
> met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw -amodel
OFCL
> -out_line_type CTC
>
> Here's the 2x2 contingency table counts:
>
> TOTAL FY_OY FY_ON FN_OY FN_ON: 60 1 0 21 38
>
> Out of 60 points, OFCL got the RI correct once, it missed it 21
times, and
> 38 times no RI was forecast or occurring.
>
> (5) Now rerun the same job for ALL models present using the "-by
AMODEL"
> option:
>
> met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw -by
AMODEL
> -out_line_type CTC -out al152017_rirw_ctc.out
>
> And write the results to an output file. You'll see that we now
have RI
> contingency table counts computed separately for each mode.
>
> There's many more options here you could play with:
>
> - Set -out_line_type CTC,CTS,MPR to dump out the contingency table
counts
> (CTC), statistics (CTS), and very detailed information for each time
point.
>
> - Specify more columns for stratifying the data... add "-by LEAD" to
break
> results down by lead time.
>
> - Change the definition of RI that you'd like to apply. Perhaps,
OFCL
> would score better when defining RI as a 25kt increase instead of
30?
>
> Hope this example helps get you going using the MET-TC tools to
quantify
> RI.
>
> Please let us know what other questions you have.
>
> Thanks,
> John Halley Gotway
> On Thu, Sep 21, 2017 at 8:50 AM, Matthew Onderlinde - NOAA Federal
via RT <
> met_help at ucar.edu> wrote:
>
> >
> > Thu Sep 21 08:50:25 2017: Request 82076 was acted upon.
> > Transaction: Ticket created by matthew.onderlinde at noaa.gov
> > Queue: met_help
> > Subject: ERROR : AsciiHeader::read() -> trouble reading file
> > Owner: Nobody
> > Requestors: matthew.onderlinde at noaa.gov
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
> >
> >
> > Hello,
> >
> > I'm trying to use MET-TC to do some rapid intensification
verification.
> > Here is the command I'm issuing:
> >
> > $> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
> >
> > which returns:
> >
> > DEBUG 1: [Source 1] Lookin: test.dat
> > DEBUG 1: Config File Default:
> > /usr/local/share/met/config/TCStatConfig_default
> > DEBUG 1: Config File User: TCStat_DTOPS_Config
> > DEBUG 2:
> > DEBUG 2: Processing Job 1: -job rirw -amodel DTOP -amodel DTOP
-bmodel
> BEST
> > -storm_id AL152017 -basin AL -cyclone 15 -rirw_time 240000
-rirw_exact
> true
> > -rirw_thresh >=30.0 -rirw_window 000000 000000 -out_line_type CTC
> > -out_line_type CTS -out_alpha 0.05000
> > ERROR :
> > ERROR : AsciiHeader::read() -> trouble reading file:
> > ERROR :
/usr/local/share/met/table_files/met_header_columns_AL,.txt
> > ERROR :
> >
> > My config file (TCStat_DTOPS_Config) and my data file (test.dat)
are
> > attached.
> >
> > It seems MET is not properly identifying the header file to use.
Could I
> > get some help troubleshooting this issue?
> >
> > To clarify, on my system, the path
/usr/local/share/met/table_files/
> > contains these met_header_columns_VX.X.txt files:
> >
> > met_header_columns_V2.0.txt met_header_columns_V3.0.txt
> > met_header_columns_V4.0.txt met_header_columns_V4.2.txt
> > met_header_columns_V5.1.txt met_header_columns_V6.0.txt
> > met_header_columns_V3.0.1.txt met_header_columns_V3.1.txt
> > met_header_columns_V4.1.txt met_header_columns_V5.0.txt
> > met_header_columns_V5.2.txt
> >
> > Thanks,
> >
> > Matt
> >
> >
> > *Matthew Onderlinde*Meteorologist/Programmer
> > NHC/Technology & Science Branch
> > (o) 305.229.4466 <(305)%20229-4466>
> > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > Signature-Block-for-Gmail.html#>
> > Matthew.Onderlinde at noaa.gov
> >
> >
>
>
------------------------------------------------
Subject: ERROR : AsciiHeader::read() -> trouble reading file
From: John Halley Gotway
Time: Thu Sep 21 12:07:48 2017
Matt,
Ah yes, sorry for the confusion. It's been a while since I've
processed
any hurricane data.
The example I sent was using ATCF track data to identify RI events in
both
the ADECK and BDECK and do categorical verification.
Yes, we added support for EDECK RI probability verification in met-
6.0:
http://www.dtcenter.org/met/users/support/release_notes/
METv6.0_release_notes.php
Using the "test.dat" data you sent:
(1) Split up "test.dat" back into edeck and bdeck files:
egrep BEST test.dat > bal152017.dat
egrep -v BEST test.dat > eal152017.dat
(2) Run tc_pairs to compute pairs:
met-6.0/bin/tc_pairs -edeck eal152017.dat -bdeck bal152017.dat
-config
TCPairsConfig -out maria_probri -v 4
(3) And then you'd run the tc_stat "-job probrirw" to compute
probabilistic
statistics, described on page 288 of the MET User's Guide:
http://www.dtcenter.org/met/users/docs/users_guide/MET_Users_Guide_v6.0.pdf
You'll notice though that you get no output from step (2). That's
because
the event hasn't occurred yet. The probability forecast was issued at
2017092106 <(201)%20709-2106> for 12, 24, 36, 48, and 72 hours in the
future. Since those times haven't occurred yet, the BEST track
doesn't
contain any verifying information and you get no output.
Try verifying the EDECK RI probabilities from the last few days.
Those
should produce data.
Thanks,
John
On Thu, Sep 21, 2017 at 11:08 AM, Matthew Onderlinde - NOAA Federal
via RT <
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
>
> Hi John,
>
> Thanks so much for taking the time to help me out. I appreciate
your
> detailed response. I've been asked to do verification of RI
probability
> forecasts. So my test.dat file contained an e-deck formatted
probability
> forecast and the best track data. What probably wasn't acceptable to
MET
> was catting the e-deck and the best track into one file. Regardless
of
> that, I get the impression that MET isn't set up to do quantitative
> verification of RI probability forecasts. Can MET produce brier
scores for
> example? If not, I may just need to write the code myself.
>
> Thanks again for your help,
>
> Matt
>
>
>
> *Matthew Onderlinde*Meteorologist/Programmer
> NHC/Technology & Science Branch
> (o) 305.229.4466
> <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> Signature-Block-for-Gmail.html#>
> Matthew.Onderlinde at noaa.gov
>
> On Thu, Sep 21, 2017 at 4:53 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Matt,
> >
> > Thanks for sending some sample data. That makes it a lot easier
to
> debug.
> >
> > The test.dat file you sent contains track data in ATCF format.
While the
> > tc_pairs tools is designed to read that data, the tc_stat tool is
not.
> > tc_stat reads the ASCII output of tc_pairs.
> >
> > Looking at the test.dat file, I'm guessing that you're trying to
compare
> > the DTOP track to the BEST track, looking for instances of rapid
> > intensification. If that's not the case, please let me know.
> >
> > But looking at the DTOP track data, I'm concerned about the 4th
column
> with
> > RI in it, and columns 9 and 10 don't have wind speeds and pressure
values
> > in the range I'd expect. My guess is that these RI lines contain
24-hour
> > differences, but the logic in MET hasn't been set up to process
this type
> > of data.
> >
> > Instead, MET expects to process raw ATCF track data. Let me work
through
> > an example...
> >
> > (1) Use the BEST track for Maria that you sent.
> >
> > grep BEST test.dat > bal152017.dat
> >
> > (2) Pull the ADECK tracks for Maria:
> >
> > wget http://ftp.nhc.noaa.gov/atcf/aid_public/aal152017.dat.gz
> >
> > gunzip aal152017.dat.gz
> > (3) Run the MET tc_pairs tool to match up the forecast/best
tracks:
> >
> > met-6.0/bin/tc_pairs -adeck aal152017.dat -bdeck bal152017.dat
-config
> > TCPairsConfig -out al152017_tcpairs
> >
> > Running using the default TCPairs configuration file results in
2,306
> > forecast tracks being processed.
> >
> > (4) Run the MET tc_stat tool to look for RI events but only for
the OFCL
> > forecast:
> >
> > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw
-amodel OFCL
> > -out_line_type CTC
> >
> > Here's the 2x2 contingency table counts:
> >
> > TOTAL FY_OY FY_ON FN_OY FN_ON: 60 1 0 21 38
> >
> > Out of 60 points, OFCL got the RI correct once, it missed it 21
times,
> and
> > 38 times no RI was forecast or occurring.
> >
> > (5) Now rerun the same job for ALL models present using the "-by
AMODEL"
> > option:
> >
> > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw -by
AMODEL
> > -out_line_type CTC -out al152017_rirw_ctc.out
> >
> > And write the results to an output file. You'll see that we now
have RI
> > contingency table counts computed separately for each mode.
> >
> > There's many more options here you could play with:
> >
> > - Set -out_line_type CTC,CTS,MPR to dump out the contingency table
counts
> > (CTC), statistics (CTS), and very detailed information for each
time
> point.
> >
> > - Specify more columns for stratifying the data... add "-by LEAD"
to
> break
> > results down by lead time.
> >
> > - Change the definition of RI that you'd like to apply. Perhaps,
OFCL
> > would score better when defining RI as a 25kt increase instead of
30?
> >
> > Hope this example helps get you going using the MET-TC tools to
quantify
> > RI.
> >
> > Please let us know what other questions you have.
> >
> > Thanks,
> > John Halley Gotway
> > On Thu, Sep 21, 2017 at 8:50 AM, Matthew Onderlinde - NOAA Federal
via
> RT <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > Thu Sep 21 08:50:25 2017: Request 82076 was acted upon.
> > > Transaction: Ticket created by matthew.onderlinde at noaa.gov
> > > Queue: met_help
> > > Subject: ERROR : AsciiHeader::read() -> trouble reading
file
> > > Owner: Nobody
> > > Requestors: matthew.onderlinde at noaa.gov
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076
> >
> > >
> > >
> > > Hello,
> > >
> > > I'm trying to use MET-TC to do some rapid intensification
verification.
> > > Here is the command I'm issuing:
> > >
> > > $> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
> > >
> > > which returns:
> > >
> > > DEBUG 1: [Source 1] Lookin: test.dat
> > > DEBUG 1: Config File Default:
> > > /usr/local/share/met/config/TCStatConfig_default
> > > DEBUG 1: Config File User: TCStat_DTOPS_Config
> > > DEBUG 2:
> > > DEBUG 2: Processing Job 1: -job rirw -amodel DTOP -amodel DTOP
-bmodel
> > BEST
> > > -storm_id AL152017 -basin AL -cyclone 15 -rirw_time 240000
-rirw_exact
> > true
> > > -rirw_thresh >=30.0 -rirw_window 000000 000000 -out_line_type
CTC
> > > -out_line_type CTS -out_alpha 0.05000
> > > ERROR :
> > > ERROR : AsciiHeader::read() -> trouble reading file:
> > > ERROR :
/usr/local/share/met/table_files/met_header_columns_AL,.txt
> > > ERROR :
> > >
> > > My config file (TCStat_DTOPS_Config) and my data file (test.dat)
are
> > > attached.
> > >
> > > It seems MET is not properly identifying the header file to use.
> Could I
> > > get some help troubleshooting this issue?
> > >
> > > To clarify, on my system, the path
/usr/local/share/met/table_files/
> > > contains these met_header_columns_VX.X.txt files:
> > >
> > > met_header_columns_V2.0.txt met_header_columns_V3.0.txt
> > > met_header_columns_V4.0.txt met_header_columns_V4.2.txt
> > > met_header_columns_V5.1.txt met_header_columns_V6.0.txt
> > > met_header_columns_V3.0.1.txt met_header_columns_V3.1.txt
> > > met_header_columns_V4.1.txt met_header_columns_V5.0.txt
> > > met_header_columns_V5.2.txt
> > >
> > > Thanks,
> > >
> > > Matt
> > >
> > >
> > > *Matthew Onderlinde*Meteorologist/Programmer
> > > NHC/Technology & Science Branch
> > > (o) 305.229.4466 <(305)%20229-4466>
> > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > Signature-Block-for-Gmail.html#>
> > > Matthew.Onderlinde at noaa.gov
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: ERROR : AsciiHeader::read() -> trouble reading file
From: Matthew Onderlinde - NOAA Federal
Time: Thu Sep 21 12:23:42 2017
Thanks John,
I issued this command:
$> tc_pairs -edeck 17091812AL1517_dtops.edk -bdeck bal152017.dat
-config
TCPairsConfig -out maria_probri -v 4
which returned:
DEBUG 1: [Source 1 of 1] EDECK Source: 17091812AL1517_dtops.edk, Model
Suffix: (nul)
DEBUG 1: [Source 1 of 1] BDECK Source: bal152017.dat, Model Suffix:
(nul)
DEBUG 1: Config File Default:
/usr/local/share/met/config/TCPairsConfig_default
DEBUG 1: Config File User: TCPairsConfig
DEBUG 1: Distance to land file: /usr/local/share/met/tc_data/
dland_global_tenth_degree.nc
DEBUG 4: Latitude/Longitude Grid Data:
DEBUG 4: lat_ll = -90
DEBUG 4: lon_ll = 180
DEBUG 4: delta_lat = 0.1
DEBUG 4: delta_lon = 0.1
DEBUG 4: Nlat = 1801
DEBUG 4: Nlon = 3601
Segmentation fault (core dumped)
The input files are attached.
Thanks,
Matt
*Matthew Onderlinde*Meteorologist/Programmer
NHC/Technology & Science Branch
(o) 305.229.4466
<http://innovim.com/wp-content/uploads/BD-F-002a-Email-Signature-
Block-for-Gmail.html#>
Matthew.Onderlinde at noaa.gov
On Thu, Sep 21, 2017 at 6:07 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Matt,
>
> Ah yes, sorry for the confusion. It's been a while since I've
processed
> any hurricane data.
>
> The example I sent was using ATCF track data to identify RI events
in both
> the ADECK and BDECK and do categorical verification.
>
> Yes, we added support for EDECK RI probability verification in met-
6.0:
>
> http://www.dtcenter.org/met/users/support/release_notes/
> METv6.0_release_notes.php
>
> Using the "test.dat" data you sent:
>
> (1) Split up "test.dat" back into edeck and bdeck files:
>
> egrep BEST test.dat > bal152017.dat
>
> egrep -v BEST test.dat > eal152017.dat
> (2) Run tc_pairs to compute pairs:
>
> met-6.0/bin/tc_pairs -edeck eal152017.dat -bdeck bal152017.dat
-config
> TCPairsConfig -out maria_probri -v 4
>
> (3) And then you'd run the tc_stat "-job probrirw" to compute
probabilistic
> statistics, described on page 288 of the MET User's Guide:
>
> http://www.dtcenter.org/met/users/docs/users_guide/MET_
> Users_Guide_v6.0.pdf
>
> You'll notice though that you get no output from step (2). That's
because
> the event hasn't occurred yet. The probability forecast was issued
at
> 2017092106 <(201)%20709-2106> for 12, 24, 36, 48, and 72 hours in
the
> future. Since those times haven't occurred yet, the BEST track
doesn't
> contain any verifying information and you get no output.
>
> Try verifying the EDECK RI probabilities from the last few days.
Those
> should produce data.
>
> Thanks,
>
> John
>
> On Thu, Sep 21, 2017 at 11:08 AM, Matthew Onderlinde - NOAA Federal
via RT
> <
> met_help at ucar.edu> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
> >
> > Hi John,
> >
> > Thanks so much for taking the time to help me out. I appreciate
your
> > detailed response. I've been asked to do verification of RI
probability
> > forecasts. So my test.dat file contained an e-deck formatted
probability
> > forecast and the best track data. What probably wasn't acceptable
to MET
> > was catting the e-deck and the best track into one file.
Regardless of
> > that, I get the impression that MET isn't set up to do
quantitative
> > verification of RI probability forecasts. Can MET produce brier
scores
> for
> > example? If not, I may just need to write the code myself.
> >
> > Thanks again for your help,
> >
> > Matt
> >
> >
> >
> > *Matthew Onderlinde*Meteorologist/Programmer
> > NHC/Technology & Science Branch
> > (o) 305.229.4466
> > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > Signature-Block-for-Gmail.html#>
> > Matthew.Onderlinde at noaa.gov
> >
> > On Thu, Sep 21, 2017 at 4:53 PM, John Halley Gotway via RT <
> > met_help at ucar.edu> wrote:
> >
> > > Matt,
> > >
> > > Thanks for sending some sample data. That makes it a lot easier
to
> > debug.
> > >
> > > The test.dat file you sent contains track data in ATCF format.
While
> the
> > > tc_pairs tools is designed to read that data, the tc_stat tool
is not.
> > > tc_stat reads the ASCII output of tc_pairs.
> > >
> > > Looking at the test.dat file, I'm guessing that you're trying to
> compare
> > > the DTOP track to the BEST track, looking for instances of rapid
> > > intensification. If that's not the case, please let me know.
> > >
> > > But looking at the DTOP track data, I'm concerned about the 4th
column
> > with
> > > RI in it, and columns 9 and 10 don't have wind speeds and
pressure
> values
> > > in the range I'd expect. My guess is that these RI lines
contain
> 24-hour
> > > differences, but the logic in MET hasn't been set up to process
this
> type
> > > of data.
> > >
> > > Instead, MET expects to process raw ATCF track data. Let me
work
> through
> > > an example...
> > >
> > > (1) Use the BEST track for Maria that you sent.
> > >
> > > grep BEST test.dat > bal152017.dat
> > >
> > > (2) Pull the ADECK tracks for Maria:
> > >
> > > wget http://ftp.nhc.noaa.gov/atcf/aid_public/aal152017.dat.gz
> > >
> > > gunzip aal152017.dat.gz
> > > (3) Run the MET tc_pairs tool to match up the forecast/best
tracks:
> > >
> > > met-6.0/bin/tc_pairs -adeck aal152017.dat -bdeck
bal152017.dat
> -config
> > > TCPairsConfig -out al152017_tcpairs
> > >
> > > Running using the default TCPairs configuration file results in
2,306
> > > forecast tracks being processed.
> > >
> > > (4) Run the MET tc_stat tool to look for RI events but only for
the
> OFCL
> > > forecast:
> > >
> > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw
-amodel
> OFCL
> > > -out_line_type CTC
> > >
> > > Here's the 2x2 contingency table counts:
> > >
> > > TOTAL FY_OY FY_ON FN_OY FN_ON: 60 1 0 21 38
> > >
> > > Out of 60 points, OFCL got the RI correct once, it missed it 21
times,
> > and
> > > 38 times no RI was forecast or occurring.
> > >
> > > (5) Now rerun the same job for ALL models present using the "-by
> AMODEL"
> > > option:
> > >
> > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw -by
AMODEL
> > > -out_line_type CTC -out al152017_rirw_ctc.out
> > >
> > > And write the results to an output file. You'll see that we now
have
> RI
> > > contingency table counts computed separately for each mode.
> > >
> > > There's many more options here you could play with:
> > >
> > > - Set -out_line_type CTC,CTS,MPR to dump out the contingency
table
> counts
> > > (CTC), statistics (CTS), and very detailed information for each
time
> > point.
> > >
> > > - Specify more columns for stratifying the data... add "-by
LEAD" to
> > break
> > > results down by lead time.
> > >
> > > - Change the definition of RI that you'd like to apply.
Perhaps, OFCL
> > > would score better when defining RI as a 25kt increase instead
of 30?
> > >
> > > Hope this example helps get you going using the MET-TC tools to
> quantify
> > > RI.
> > >
> > > Please let us know what other questions you have.
> > >
> > > Thanks,
> > > John Halley Gotway
> > > On Thu, Sep 21, 2017 at 8:50 AM, Matthew Onderlinde - NOAA
Federal via
> > RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > Thu Sep 21 08:50:25 2017: Request 82076 was acted upon.
> > > > Transaction: Ticket created by matthew.onderlinde at noaa.gov
> > > > Queue: met_help
> > > > Subject: ERROR : AsciiHeader::read() -> trouble reading
file
> > > > Owner: Nobody
> > > > Requestors: matthew.onderlinde at noaa.gov
> > > > Status: new
> > > > Ticket <URL: https://rt.rap.ucar.edu/rt/
> Ticket/Display.html?id=82076
> > >
> > > >
> > > >
> > > > Hello,
> > > >
> > > > I'm trying to use MET-TC to do some rapid intensification
> verification.
> > > > Here is the command I'm issuing:
> > > >
> > > > $> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
> > > >
> > > > which returns:
> > > >
> > > > DEBUG 1: [Source 1] Lookin: test.dat
> > > > DEBUG 1: Config File Default:
> > > > /usr/local/share/met/config/TCStatConfig_default
> > > > DEBUG 1: Config File User: TCStat_DTOPS_Config
> > > > DEBUG 2:
> > > > DEBUG 2: Processing Job 1: -job rirw -amodel DTOP -amodel DTOP
> -bmodel
> > > BEST
> > > > -storm_id AL152017 -basin AL -cyclone 15 -rirw_time 240000
> -rirw_exact
> > > true
> > > > -rirw_thresh >=30.0 -rirw_window 000000 000000 -out_line_type
CTC
> > > > -out_line_type CTS -out_alpha 0.05000
> > > > ERROR :
> > > > ERROR : AsciiHeader::read() -> trouble reading file:
> > > > ERROR :
/usr/local/share/met/table_files/met_header_columns_AL,.txt
> > > > ERROR :
> > > >
> > > > My config file (TCStat_DTOPS_Config) and my data file
(test.dat) are
> > > > attached.
> > > >
> > > > It seems MET is not properly identifying the header file to
use.
> > Could I
> > > > get some help troubleshooting this issue?
> > > >
> > > > To clarify, on my system, the path
/usr/local/share/met/table_files/
> > > > contains these met_header_columns_VX.X.txt files:
> > > >
> > > > met_header_columns_V2.0.txt met_header_columns_V3.0.txt
> > > > met_header_columns_V4.0.txt met_header_columns_V4.2.txt
> > > > met_header_columns_V5.1.txt met_header_columns_V6.0.txt
> > > > met_header_columns_V3.0.1.txt met_header_columns_V3.1.txt
> > > > met_header_columns_V4.1.txt met_header_columns_V5.0.txt
> > > > met_header_columns_V5.2.txt
> > > >
> > > > Thanks,
> > > >
> > > > Matt
> > > >
> > > >
> > > > *Matthew Onderlinde*Meteorologist/Programmer
> > > > NHC/Technology & Science Branch
> > > > (o) 305.229.4466 <(305)%20229-4466>
> > > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > > Signature-Block-for-Gmail.html#>
> > > > Matthew.Onderlinde at noaa.gov
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: ERROR : AsciiHeader::read() -> trouble reading file
From: John Halley Gotway
Time: Thu Sep 21 12:51:26 2017
Matt,
Thanks again for sending the sample data. When I run met-6.0 using
all the
latest patches, I do not get the segfault you're seeing. But tc_pairs
finds 0 EDECK tracks:
DEBUG 2: Processing 1 EDECK file(s).
DEBUG 2: Filtering 0 probabilities based on config file settings.
DEBUG 2: Matching 0 EDECK probabilities to 1 BDECK tracks.
Looking closely in your config file, I see why:
model = ["DTOPS"];
You meant to say this:
model = ["DTOP"];
But actually, you don't need to specify anything there. If it's an
empty
list, MET will just process all the data you pass to it, regardless of
the
model name:
model = [ ];
I did try running met-6.0 as originally released with no patches, and
that
had the same behavior. No segfault.
So I'm not sure what's going on there.
John
On Thu, Sep 21, 2017 at 12:23 PM, Matthew Onderlinde - NOAA Federal
via RT <
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
>
> Thanks John,
>
> I issued this command:
>
> $> tc_pairs -edeck 17091812AL1517_dtops.edk -bdeck bal152017.dat
-config
> TCPairsConfig -out maria_probri -v 4
>
> which returned:
>
> DEBUG 1: [Source 1 of 1] EDECK Source: 17091812AL1517_dtops.edk,
Model
> Suffix: (nul)
> DEBUG 1: [Source 1 of 1] BDECK Source: bal152017.dat, Model Suffix:
(nul)
> DEBUG 1: Config File Default:
> /usr/local/share/met/config/TCPairsConfig_default
> DEBUG 1: Config File User: TCPairsConfig
> DEBUG 1: Distance to land file: /usr/local/share/met/tc_data/
> dland_global_tenth_degree.nc
> DEBUG 4: Latitude/Longitude Grid Data:
> DEBUG 4: lat_ll = -90
> DEBUG 4: lon_ll = 180
> DEBUG 4: delta_lat = 0.1
> DEBUG 4: delta_lon = 0.1
> DEBUG 4: Nlat = 1801
> DEBUG 4: Nlon = 3601
> Segmentation fault (core dumped)
>
> The input files are attached.
>
> Thanks,
>
> Matt
>
>
>
>
> *Matthew Onderlinde*Meteorologist/Programmer
> NHC/Technology & Science Branch
> (o) 305.229.4466
> <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> Signature-Block-for-Gmail.html#>
> Matthew.Onderlinde at noaa.gov
>
> On Thu, Sep 21, 2017 at 6:07 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Matt,
> >
> > Ah yes, sorry for the confusion. It's been a while since I've
processed
> > any hurricane data.
> >
> > The example I sent was using ATCF track data to identify RI events
in
> both
> > the ADECK and BDECK and do categorical verification.
> >
> > Yes, we added support for EDECK RI probability verification in
met-6.0:
> >
> > http://www.dtcenter.org/met/users/support/release_notes/
> > METv6.0_release_notes.php
> >
> > Using the "test.dat" data you sent:
> >
> > (1) Split up "test.dat" back into edeck and bdeck files:
> >
> > egrep BEST test.dat > bal152017.dat
> >
> > egrep -v BEST test.dat > eal152017.dat
> > (2) Run tc_pairs to compute pairs:
> >
> > met-6.0/bin/tc_pairs -edeck eal152017.dat -bdeck bal152017.dat
-config
> > TCPairsConfig -out maria_probri -v 4
> >
> > (3) And then you'd run the tc_stat "-job probrirw" to compute
> probabilistic
> > statistics, described on page 288 of the MET User's Guide:
> >
> > http://www.dtcenter.org/met/users/docs/users_guide/MET_
> > Users_Guide_v6.0.pdf
> >
> > You'll notice though that you get no output from step (2). That's
> because
> > the event hasn't occurred yet. The probability forecast was
issued at
> > 2017092106 <(201)%20709-2106> for 12, 24, 36, 48, and 72 hours in
the
> > future. Since those times haven't occurred yet, the BEST track
doesn't
> > contain any verifying information and you get no output.
> >
> > Try verifying the EDECK RI probabilities from the last few days.
Those
> > should produce data.
> >
> > Thanks,
> >
> > John
> >
> > On Thu, Sep 21, 2017 at 11:08 AM, Matthew Onderlinde - NOAA
Federal via
> RT
> > <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
> > >
> > > Hi John,
> > >
> > > Thanks so much for taking the time to help me out. I appreciate
your
> > > detailed response. I've been asked to do verification of RI
> probability
> > > forecasts. So my test.dat file contained an e-deck formatted
> probability
> > > forecast and the best track data. What probably wasn't
acceptable to
> MET
> > > was catting the e-deck and the best track into one file.
Regardless of
> > > that, I get the impression that MET isn't set up to do
quantitative
> > > verification of RI probability forecasts. Can MET produce brier
scores
> > for
> > > example? If not, I may just need to write the code myself.
> > >
> > > Thanks again for your help,
> > >
> > > Matt
> > >
> > >
> > >
> > > *Matthew Onderlinde*Meteorologist/Programmer
> > > NHC/Technology & Science Branch
> > > (o) 305.229.4466
> > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > Signature-Block-for-Gmail.html#>
> > > Matthew.Onderlinde at noaa.gov
> > >
> > > On Thu, Sep 21, 2017 at 4:53 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > > Matt,
> > > >
> > > > Thanks for sending some sample data. That makes it a lot
easier to
> > > debug.
> > > >
> > > > The test.dat file you sent contains track data in ATCF format.
While
> > the
> > > > tc_pairs tools is designed to read that data, the tc_stat tool
is
> not.
> > > > tc_stat reads the ASCII output of tc_pairs.
> > > >
> > > > Looking at the test.dat file, I'm guessing that you're trying
to
> > compare
> > > > the DTOP track to the BEST track, looking for instances of
rapid
> > > > intensification. If that's not the case, please let me know.
> > > >
> > > > But looking at the DTOP track data, I'm concerned about the
4th
> column
> > > with
> > > > RI in it, and columns 9 and 10 don't have wind speeds and
pressure
> > values
> > > > in the range I'd expect. My guess is that these RI lines
contain
> > 24-hour
> > > > differences, but the logic in MET hasn't been set up to
process this
> > type
> > > > of data.
> > > >
> > > > Instead, MET expects to process raw ATCF track data. Let me
work
> > through
> > > > an example...
> > > >
> > > > (1) Use the BEST track for Maria that you sent.
> > > >
> > > > grep BEST test.dat > bal152017.dat
> > > >
> > > > (2) Pull the ADECK tracks for Maria:
> > > >
> > > > wget
http://ftp.nhc.noaa.gov/atcf/aid_public/aal152017.dat.gz
> > > >
> > > > gunzip aal152017.dat.gz
> > > > (3) Run the MET tc_pairs tool to match up the forecast/best
tracks:
> > > >
> > > > met-6.0/bin/tc_pairs -adeck aal152017.dat -bdeck
bal152017.dat
> > -config
> > > > TCPairsConfig -out al152017_tcpairs
> > > >
> > > > Running using the default TCPairs configuration file results
in 2,306
> > > > forecast tracks being processed.
> > > >
> > > > (4) Run the MET tc_stat tool to look for RI events but only
for the
> > OFCL
> > > > forecast:
> > > >
> > > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw
-amodel
> > OFCL
> > > > -out_line_type CTC
> > > >
> > > > Here's the 2x2 contingency table counts:
> > > >
> > > > TOTAL FY_OY FY_ON FN_OY FN_ON: 60 1 0 21 38
> > > >
> > > > Out of 60 points, OFCL got the RI correct once, it missed it
21
> times,
> > > and
> > > > 38 times no RI was forecast or occurring.
> > > >
> > > > (5) Now rerun the same job for ALL models present using the "-
by
> > AMODEL"
> > > > option:
> > > >
> > > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw
-by
> AMODEL
> > > > -out_line_type CTC -out al152017_rirw_ctc.out
> > > >
> > > > And write the results to an output file. You'll see that we
now have
> > RI
> > > > contingency table counts computed separately for each mode.
> > > >
> > > > There's many more options here you could play with:
> > > >
> > > > - Set -out_line_type CTC,CTS,MPR to dump out the contingency
table
> > counts
> > > > (CTC), statistics (CTS), and very detailed information for
each time
> > > point.
> > > >
> > > > - Specify more columns for stratifying the data... add "-by
LEAD" to
> > > break
> > > > results down by lead time.
> > > >
> > > > - Change the definition of RI that you'd like to apply.
Perhaps,
> OFCL
> > > > would score better when defining RI as a 25kt increase instead
of 30?
> > > >
> > > > Hope this example helps get you going using the MET-TC tools
to
> > quantify
> > > > RI.
> > > >
> > > > Please let us know what other questions you have.
> > > >
> > > > Thanks,
> > > > John Halley Gotway
> > > > On Thu, Sep 21, 2017 at 8:50 AM, Matthew Onderlinde - NOAA
Federal
> via
> > > RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > Thu Sep 21 08:50:25 2017: Request 82076 was acted upon.
> > > > > Transaction: Ticket created by matthew.onderlinde at noaa.gov
> > > > > Queue: met_help
> > > > > Subject: ERROR : AsciiHeader::read() -> trouble reading
file
> > > > > Owner: Nobody
> > > > > Requestors: matthew.onderlinde at noaa.gov
> > > > > Status: new
> > > > > Ticket <URL: https://rt.rap.ucar.edu/rt/
> > Ticket/Display.html?id=82076
> > > >
> > > > >
> > > > >
> > > > > Hello,
> > > > >
> > > > > I'm trying to use MET-TC to do some rapid intensification
> > verification.
> > > > > Here is the command I'm issuing:
> > > > >
> > > > > $> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
> > > > >
> > > > > which returns:
> > > > >
> > > > > DEBUG 1: [Source 1] Lookin: test.dat
> > > > > DEBUG 1: Config File Default:
> > > > > /usr/local/share/met/config/TCStatConfig_default
> > > > > DEBUG 1: Config File User: TCStat_DTOPS_Config
> > > > > DEBUG 2:
> > > > > DEBUG 2: Processing Job 1: -job rirw -amodel DTOP -amodel
DTOP
> > -bmodel
> > > > BEST
> > > > > -storm_id AL152017 -basin AL -cyclone 15 -rirw_time 240000
> > -rirw_exact
> > > > true
> > > > > -rirw_thresh >=30.0 -rirw_window 000000 000000
-out_line_type CTC
> > > > > -out_line_type CTS -out_alpha 0.05000
> > > > > ERROR :
> > > > > ERROR : AsciiHeader::read() -> trouble reading file:
> > > > > ERROR : /usr/local/share/met/table_fil
> es/met_header_columns_AL,.txt
> > > > > ERROR :
> > > > >
> > > > > My config file (TCStat_DTOPS_Config) and my data file
(test.dat)
> are
> > > > > attached.
> > > > >
> > > > > It seems MET is not properly identifying the header file to
use.
> > > Could I
> > > > > get some help troubleshooting this issue?
> > > > >
> > > > > To clarify, on my system, the path
/usr/local/share/met/table_fil
> es/
> > > > > contains these met_header_columns_VX.X.txt files:
> > > > >
> > > > > met_header_columns_V2.0.txt met_header_columns_V3.0.txt
> > > > > met_header_columns_V4.0.txt met_header_columns_V4.2.txt
> > > > > met_header_columns_V5.1.txt met_header_columns_V6.0.txt
> > > > > met_header_columns_V3.0.1.txt met_header_columns_V3.1.txt
> > > > > met_header_columns_V4.1.txt met_header_columns_V5.0.txt
> > > > > met_header_columns_V5.2.txt
> > > > >
> > > > > Thanks,
> > > > >
> > > > > Matt
> > > > >
> > > > >
> > > > > *Matthew Onderlinde*Meteorologist/Programmer
> > > > > NHC/Technology & Science Branch
> > > > > (o) 305.229.4466 <(305)%20229-4466> <(305)%20229-4466>
> > > > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > > > Signature-Block-for-Gmail.html#>
> > > > > Matthew.Onderlinde at noaa.gov
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: ERROR : AsciiHeader::read() -> trouble reading file
From: Matthew Onderlinde - NOAA Federal
Time: Thu Sep 21 13:01:33 2017
Okay, thanks John. I do get the seg fault when I correct DTOPS to
DTOP, or
when I leave the model config option blank. It must be an issue with
my
MET installation. I will look into this with our tech support.
Thanks,
Matt
*Matthew Onderlinde*Meteorologist/Programmer
NHC/Technology & Science Branch
(o) 305.229.4466
<http://innovim.com/wp-content/uploads/BD-F-002a-Email-Signature-
Block-for-Gmail.html#>
Matthew.Onderlinde at noaa.gov
On Thu, Sep 21, 2017 at 6:51 PM, John Halley Gotway via RT <
met_help at ucar.edu> wrote:
> Matt,
>
> Thanks again for sending the sample data. When I run met-6.0 using
all the
> latest patches, I do not get the segfault you're seeing. But
tc_pairs
> finds 0 EDECK tracks:
>
> DEBUG 2: Processing 1 EDECK file(s).
>
> DEBUG 2: Filtering 0 probabilities based on config file settings.
>
> DEBUG 2: Matching 0 EDECK probabilities to 1 BDECK tracks.
>
> Looking closely in your config file, I see why:
>
> model = ["DTOPS"];
>
> You meant to say this:
>
> model = ["DTOP"];
>
> But actually, you don't need to specify anything there. If it's an
empty
> list, MET will just process all the data you pass to it, regardless
of the
> model name:
>
> model = [ ];
>
> I did try running met-6.0 as originally released with no patches,
and that
> had the same behavior. No segfault.
>
> So I'm not sure what's going on there.
>
> John
>
> On Thu, Sep 21, 2017 at 12:23 PM, Matthew Onderlinde - NOAA Federal
via RT
> <
> met_help at ucar.edu> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
> >
> > Thanks John,
> >
> > I issued this command:
> >
> > $> tc_pairs -edeck 17091812AL1517_dtops.edk -bdeck bal152017.dat
-config
> > TCPairsConfig -out maria_probri -v 4
> >
> > which returned:
> >
> > DEBUG 1: [Source 1 of 1] EDECK Source: 17091812AL1517_dtops.edk,
Model
> > Suffix: (nul)
> > DEBUG 1: [Source 1 of 1] BDECK Source: bal152017.dat, Model
Suffix: (nul)
> > DEBUG 1: Config File Default:
> > /usr/local/share/met/config/TCPairsConfig_default
> > DEBUG 1: Config File User: TCPairsConfig
> > DEBUG 1: Distance to land file: /usr/local/share/met/tc_data/
> > dland_global_tenth_degree.nc
> > DEBUG 4: Latitude/Longitude Grid Data:
> > DEBUG 4: lat_ll = -90
> > DEBUG 4: lon_ll = 180
> > DEBUG 4: delta_lat = 0.1
> > DEBUG 4: delta_lon = 0.1
> > DEBUG 4: Nlat = 1801
> > DEBUG 4: Nlon = 3601
> > Segmentation fault (core dumped)
> >
> > The input files are attached.
> >
> > Thanks,
> >
> > Matt
> >
> >
> >
> >
> > *Matthew Onderlinde*Meteorologist/Programmer
> > NHC/Technology & Science Branch
> > (o) 305.229.4466
> > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > Signature-Block-for-Gmail.html#>
> > Matthew.Onderlinde at noaa.gov
> >
> > On Thu, Sep 21, 2017 at 6:07 PM, John Halley Gotway via RT <
> > met_help at ucar.edu> wrote:
> >
> > > Matt,
> > >
> > > Ah yes, sorry for the confusion. It's been a while since I've
> processed
> > > any hurricane data.
> > >
> > > The example I sent was using ATCF track data to identify RI
events in
> > both
> > > the ADECK and BDECK and do categorical verification.
> > >
> > > Yes, we added support for EDECK RI probability verification in
met-6.0:
> > >
> > > http://www.dtcenter.org/met/users/support/release_notes/
> > > METv6.0_release_notes.php
> > >
> > > Using the "test.dat" data you sent:
> > >
> > > (1) Split up "test.dat" back into edeck and bdeck files:
> > >
> > > egrep BEST test.dat > bal152017.dat
> > >
> > > egrep -v BEST test.dat > eal152017.dat
> > > (2) Run tc_pairs to compute pairs:
> > >
> > > met-6.0/bin/tc_pairs -edeck eal152017.dat -bdeck
bal152017.dat
> -config
> > > TCPairsConfig -out maria_probri -v 4
> > >
> > > (3) And then you'd run the tc_stat "-job probrirw" to compute
> > probabilistic
> > > statistics, described on page 288 of the MET User's Guide:
> > >
> > > http://www.dtcenter.org/met/users/docs/users_guide/MET_
> > > Users_Guide_v6.0.pdf
> > >
> > > You'll notice though that you get no output from step (2).
That's
> > because
> > > the event hasn't occurred yet. The probability forecast was
issued at
> > > 2017092106 <(201)%20709-2106> for 12, 24, 36, 48, and 72 hours
in the
> > > future. Since those times haven't occurred yet, the BEST track
doesn't
> > > contain any verifying information and you get no output.
> > >
> > > Try verifying the EDECK RI probabilities from the last few days.
Those
> > > should produce data.
> > >
> > > Thanks,
> > >
> > > John
> > >
> > > On Thu, Sep 21, 2017 at 11:08 AM, Matthew Onderlinde - NOAA
Federal via
> > RT
> > > <
> > > met_help at ucar.edu> wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076
>
> > > >
> > > > Hi John,
> > > >
> > > > Thanks so much for taking the time to help me out. I
appreciate your
> > > > detailed response. I've been asked to do verification of RI
> > probability
> > > > forecasts. So my test.dat file contained an e-deck formatted
> > probability
> > > > forecast and the best track data. What probably wasn't
acceptable to
> > MET
> > > > was catting the e-deck and the best track into one file.
Regardless
> of
> > > > that, I get the impression that MET isn't set up to do
quantitative
> > > > verification of RI probability forecasts. Can MET produce
brier
> scores
> > > for
> > > > example? If not, I may just need to write the code myself.
> > > >
> > > > Thanks again for your help,
> > > >
> > > > Matt
> > > >
> > > >
> > > >
> > > > *Matthew Onderlinde*Meteorologist/Programmer
> > > > NHC/Technology & Science Branch
> > > > (o) 305.229.4466
> > > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > > Signature-Block-for-Gmail.html#>
> > > > Matthew.Onderlinde at noaa.gov
> > > >
> > > > On Thu, Sep 21, 2017 at 4:53 PM, John Halley Gotway via RT <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > > Matt,
> > > > >
> > > > > Thanks for sending some sample data. That makes it a lot
easier to
> > > > debug.
> > > > >
> > > > > The test.dat file you sent contains track data in ATCF
format.
> While
> > > the
> > > > > tc_pairs tools is designed to read that data, the tc_stat
tool is
> > not.
> > > > > tc_stat reads the ASCII output of tc_pairs.
> > > > >
> > > > > Looking at the test.dat file, I'm guessing that you're
trying to
> > > compare
> > > > > the DTOP track to the BEST track, looking for instances of
rapid
> > > > > intensification. If that's not the case, please let me
know.
> > > > >
> > > > > But looking at the DTOP track data, I'm concerned about the
4th
> > column
> > > > with
> > > > > RI in it, and columns 9 and 10 don't have wind speeds and
pressure
> > > values
> > > > > in the range I'd expect. My guess is that these RI lines
contain
> > > 24-hour
> > > > > differences, but the logic in MET hasn't been set up to
process
> this
> > > type
> > > > > of data.
> > > > >
> > > > > Instead, MET expects to process raw ATCF track data. Let me
work
> > > through
> > > > > an example...
> > > > >
> > > > > (1) Use the BEST track for Maria that you sent.
> > > > >
> > > > > grep BEST test.dat > bal152017.dat
> > > > >
> > > > > (2) Pull the ADECK tracks for Maria:
> > > > >
> > > > > wget
http://ftp.nhc.noaa.gov/atcf/aid_public/aal152017.dat.gz
> > > > >
> > > > > gunzip aal152017.dat.gz
> > > > > (3) Run the MET tc_pairs tool to match up the forecast/best
tracks:
> > > > >
> > > > > met-6.0/bin/tc_pairs -adeck aal152017.dat -bdeck
bal152017.dat
> > > -config
> > > > > TCPairsConfig -out al152017_tcpairs
> > > > >
> > > > > Running using the default TCPairs configuration file results
in
> 2,306
> > > > > forecast tracks being processed.
> > > > >
> > > > > (4) Run the MET tc_stat tool to look for RI events but only
for the
> > > OFCL
> > > > > forecast:
> > > > >
> > > > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw
-amodel
> > > OFCL
> > > > > -out_line_type CTC
> > > > >
> > > > > Here's the 2x2 contingency table counts:
> > > > >
> > > > > TOTAL FY_OY FY_ON FN_OY FN_ON: 60 1 0 21 38
> > > > >
> > > > > Out of 60 points, OFCL got the RI correct once, it missed it
21
> > times,
> > > > and
> > > > > 38 times no RI was forecast or occurring.
> > > > >
> > > > > (5) Now rerun the same job for ALL models present using the
"-by
> > > AMODEL"
> > > > > option:
> > > > >
> > > > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job rirw
-by
> > AMODEL
> > > > > -out_line_type CTC -out al152017_rirw_ctc.out
> > > > >
> > > > > And write the results to an output file. You'll see that we
now
> have
> > > RI
> > > > > contingency table counts computed separately for each mode.
> > > > >
> > > > > There's many more options here you could play with:
> > > > >
> > > > > - Set -out_line_type CTC,CTS,MPR to dump out the contingency
table
> > > counts
> > > > > (CTC), statistics (CTS), and very detailed information for
each
> time
> > > > point.
> > > > >
> > > > > - Specify more columns for stratifying the data... add "-by
LEAD"
> to
> > > > break
> > > > > results down by lead time.
> > > > >
> > > > > - Change the definition of RI that you'd like to apply.
Perhaps,
> > OFCL
> > > > > would score better when defining RI as a 25kt increase
instead of
> 30?
> > > > >
> > > > > Hope this example helps get you going using the MET-TC tools
to
> > > quantify
> > > > > RI.
> > > > >
> > > > > Please let us know what other questions you have.
> > > > >
> > > > > Thanks,
> > > > > John Halley Gotway
> > > > > On Thu, Sep 21, 2017 at 8:50 AM, Matthew Onderlinde - NOAA
Federal
> > via
> > > > RT <
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > >
> > > > > > Thu Sep 21 08:50:25 2017: Request 82076 was acted upon.
> > > > > > Transaction: Ticket created by matthew.onderlinde at noaa.gov
> > > > > > Queue: met_help
> > > > > > Subject: ERROR : AsciiHeader::read() -> trouble
reading file
> > > > > > Owner: Nobody
> > > > > > Requestors: matthew.onderlinde at noaa.gov
> > > > > > Status: new
> > > > > > Ticket <URL: https://rt.rap.ucar.edu/rt/
> > > Ticket/Display.html?id=82076
> > > > >
> > > > > >
> > > > > >
> > > > > > Hello,
> > > > > >
> > > > > > I'm trying to use MET-TC to do some rapid intensification
> > > verification.
> > > > > > Here is the command I'm issuing:
> > > > > >
> > > > > > $> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
> > > > > >
> > > > > > which returns:
> > > > > >
> > > > > > DEBUG 1: [Source 1] Lookin: test.dat
> > > > > > DEBUG 1: Config File Default:
> > > > > > /usr/local/share/met/config/TCStatConfig_default
> > > > > > DEBUG 1: Config File User: TCStat_DTOPS_Config
> > > > > > DEBUG 2:
> > > > > > DEBUG 2: Processing Job 1: -job rirw -amodel DTOP -amodel
DTOP
> > > -bmodel
> > > > > BEST
> > > > > > -storm_id AL152017 -basin AL -cyclone 15 -rirw_time 240000
> > > -rirw_exact
> > > > > true
> > > > > > -rirw_thresh >=30.0 -rirw_window 000000 000000
-out_line_type CTC
> > > > > > -out_line_type CTS -out_alpha 0.05000
> > > > > > ERROR :
> > > > > > ERROR : AsciiHeader::read() -> trouble reading file:
> > > > > > ERROR : /usr/local/share/met/table_fil
> > es/met_header_columns_AL,.txt
> > > > > > ERROR :
> > > > > >
> > > > > > My config file (TCStat_DTOPS_Config) and my data file
(test.dat)
> > are
> > > > > > attached.
> > > > > >
> > > > > > It seems MET is not properly identifying the header file
to use.
> > > > Could I
> > > > > > get some help troubleshooting this issue?
> > > > > >
> > > > > > To clarify, on my system, the path
/usr/local/share/met/table_fil
> > es/
> > > > > > contains these met_header_columns_VX.X.txt files:
> > > > > >
> > > > > > met_header_columns_V2.0.txt met_header_columns_V3.0.txt
> > > > > > met_header_columns_V4.0.txt met_header_columns_V4.2.txt
> > > > > > met_header_columns_V5.1.txt met_header_columns_V6.0.txt
> > > > > > met_header_columns_V3.0.1.txt met_header_columns_V3.1.txt
> > > > > > met_header_columns_V4.1.txt met_header_columns_V5.0.txt
> > > > > > met_header_columns_V5.2.txt
> > > > > >
> > > > > > Thanks,
> > > > > >
> > > > > > Matt
> > > > > >
> > > > > >
> > > > > > *Matthew Onderlinde*Meteorologist/Programmer
> > > > > > NHC/Technology & Science Branch
> > > > > > (o) 305.229.4466 <(305)%20229-4466> <(305)%20229-4466>
> > > > > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > > > > Signature-Block-for-Gmail.html#>
> > > > > > Matthew.Onderlinde at noaa.gov
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
Subject: ERROR : AsciiHeader::read() -> trouble reading file
From: John Halley Gotway
Time: Thu Sep 21 14:17:51 2017
Matt,
FYI, I used the data you sent that I was able to process through
tc_pairs
to run a tc_stat job. And there I did get a segfault.
The tc_stat PROBRIRW needs to have the -dump_row option set, but I
forgot
to add a check to the code to make sure that it actually is. When
it's not
set, you get a segfault.
I'm working on a bugfix for this. And rather than requiring the user
to
actually set that option, I could instead just write the output to a
temp
file.
I'll let you know when I have a fix.
John
On Thu, Sep 21, 2017 at 1:01 PM, Matthew Onderlinde - NOAA Federal via
RT <
met_help at ucar.edu> wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
>
> Okay, thanks John. I do get the seg fault when I correct DTOPS to
DTOP, or
> when I leave the model config option blank. It must be an issue
with my
> MET installation. I will look into this with our tech support.
>
> Thanks,
>
> Matt
>
>
>
> *Matthew Onderlinde*Meteorologist/Programmer
> NHC/Technology & Science Branch
> (o) 305.229.4466
> <http://innovim.com/wp-content/uploads/BD-F-002a-
> Email-Signature-Block-for-Gmail.html#>
> Matthew.Onderlinde at noaa.gov
>
> On Thu, Sep 21, 2017 at 6:51 PM, John Halley Gotway via RT <
> met_help at ucar.edu> wrote:
>
> > Matt,
> >
> > Thanks again for sending the sample data. When I run met-6.0
using all
> the
> > latest patches, I do not get the segfault you're seeing. But
tc_pairs
> > finds 0 EDECK tracks:
> >
> > DEBUG 2: Processing 1 EDECK file(s).
> >
> > DEBUG 2: Filtering 0 probabilities based on config file settings.
> >
> > DEBUG 2: Matching 0 EDECK probabilities to 1 BDECK tracks.
> >
> > Looking closely in your config file, I see why:
> >
> > model = ["DTOPS"];
> >
> > You meant to say this:
> >
> > model = ["DTOP"];
> >
> > But actually, you don't need to specify anything there. If it's
an empty
> > list, MET will just process all the data you pass to it,
regardless of
> the
> > model name:
> >
> > model = [ ];
> >
> > I did try running met-6.0 as originally released with no patches,
and
> that
> > had the same behavior. No segfault.
> >
> > So I'm not sure what's going on there.
> >
> > John
> >
> > On Thu, Sep 21, 2017 at 12:23 PM, Matthew Onderlinde - NOAA
Federal via
> RT
> > <
> > met_help at ucar.edu> wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
> > >
> > > Thanks John,
> > >
> > > I issued this command:
> > >
> > > $> tc_pairs -edeck 17091812AL1517_dtops.edk -bdeck bal152017.dat
> -config
> > > TCPairsConfig -out maria_probri -v 4
> > >
> > > which returned:
> > >
> > > DEBUG 1: [Source 1 of 1] EDECK Source: 17091812AL1517_dtops.edk,
Model
> > > Suffix: (nul)
> > > DEBUG 1: [Source 1 of 1] BDECK Source: bal152017.dat, Model
Suffix:
> (nul)
> > > DEBUG 1: Config File Default:
> > > /usr/local/share/met/config/TCPairsConfig_default
> > > DEBUG 1: Config File User: TCPairsConfig
> > > DEBUG 1: Distance to land file: /usr/local/share/met/tc_data/
> > > dland_global_tenth_degree.nc
> > > DEBUG 4: Latitude/Longitude Grid Data:
> > > DEBUG 4: lat_ll = -90
> > > DEBUG 4: lon_ll = 180
> > > DEBUG 4: delta_lat = 0.1
> > > DEBUG 4: delta_lon = 0.1
> > > DEBUG 4: Nlat = 1801
> > > DEBUG 4: Nlon = 3601
> > > Segmentation fault (core dumped)
> > >
> > > The input files are attached.
> > >
> > > Thanks,
> > >
> > > Matt
> > >
> > >
> > >
> > >
> > > *Matthew Onderlinde*Meteorologist/Programmer
> > > NHC/Technology & Science Branch
> > > (o) 305.229.4466
> > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > Signature-Block-for-Gmail.html#>
> > > Matthew.Onderlinde at noaa.gov
> > >
> > > On Thu, Sep 21, 2017 at 6:07 PM, John Halley Gotway via RT <
> > > met_help at ucar.edu> wrote:
> > >
> > > > Matt,
> > > >
> > > > Ah yes, sorry for the confusion. It's been a while since I've
> > processed
> > > > any hurricane data.
> > > >
> > > > The example I sent was using ATCF track data to identify RI
events in
> > > both
> > > > the ADECK and BDECK and do categorical verification.
> > > >
> > > > Yes, we added support for EDECK RI probability verification in
> met-6.0:
> > > >
> > > > http://www.dtcenter.org/met/users/support/release_notes/
> > > > METv6.0_release_notes.php
> > > >
> > > > Using the "test.dat" data you sent:
> > > >
> > > > (1) Split up "test.dat" back into edeck and bdeck files:
> > > >
> > > > egrep BEST test.dat > bal152017.dat
> > > >
> > > > egrep -v BEST test.dat > eal152017.dat
> > > > (2) Run tc_pairs to compute pairs:
> > > >
> > > > met-6.0/bin/tc_pairs -edeck eal152017.dat -bdeck
bal152017.dat
> > -config
> > > > TCPairsConfig -out maria_probri -v 4
> > > >
> > > > (3) And then you'd run the tc_stat "-job probrirw" to compute
> > > probabilistic
> > > > statistics, described on page 288 of the MET User's Guide:
> > > >
> > > > http://www.dtcenter.org/met/users/docs/users_guide/MET_
> > > > Users_Guide_v6.0.pdf
> > > >
> > > > You'll notice though that you get no output from step (2).
That's
> > > because
> > > > the event hasn't occurred yet. The probability forecast was
issued
> at
> > > > 2017092106 <(201)%20709-2106> for 12, 24, 36, 48, and 72 hours
in
> the
> > > > future. Since those times haven't occurred yet, the BEST
track
> doesn't
> > > > contain any verifying information and you get no output.
> > > >
> > > > Try verifying the EDECK RI probabilities from the last few
days.
> Those
> > > > should produce data.
> > > >
> > > > Thanks,
> > > >
> > > > John
> > > >
> > > > On Thu, Sep 21, 2017 at 11:08 AM, Matthew Onderlinde - NOAA
Federal
> via
> > > RT
> > > > <
> > > > met_help at ucar.edu> wrote:
> > > >
> > > > >
> > > > > <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=82076 >
> > > > >
> > > > > Hi John,
> > > > >
> > > > > Thanks so much for taking the time to help me out. I
appreciate
> your
> > > > > detailed response. I've been asked to do verification of RI
> > > probability
> > > > > forecasts. So my test.dat file contained an e-deck
formatted
> > > probability
> > > > > forecast and the best track data. What probably wasn't
acceptable
> to
> > > MET
> > > > > was catting the e-deck and the best track into one file.
> Regardless
> > of
> > > > > that, I get the impression that MET isn't set up to do
quantitative
> > > > > verification of RI probability forecasts. Can MET produce
brier
> > scores
> > > > for
> > > > > example? If not, I may just need to write the code myself.
> > > > >
> > > > > Thanks again for your help,
> > > > >
> > > > > Matt
> > > > >
> > > > >
> > > > >
> > > > > *Matthew Onderlinde*Meteorologist/Programmer
> > > > > NHC/Technology & Science Branch
> > > > > (o) 305.229.4466
> > > > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > > > Signature-Block-for-Gmail.html#>
> > > > > Matthew.Onderlinde at noaa.gov
> > > > >
> > > > > On Thu, Sep 21, 2017 at 4:53 PM, John Halley Gotway via RT <
> > > > > met_help at ucar.edu> wrote:
> > > > >
> > > > > > Matt,
> > > > > >
> > > > > > Thanks for sending some sample data. That makes it a lot
easier
> to
> > > > > debug.
> > > > > >
> > > > > > The test.dat file you sent contains track data in ATCF
format.
> > While
> > > > the
> > > > > > tc_pairs tools is designed to read that data, the tc_stat
tool is
> > > not.
> > > > > > tc_stat reads the ASCII output of tc_pairs.
> > > > > >
> > > > > > Looking at the test.dat file, I'm guessing that you're
trying to
> > > > compare
> > > > > > the DTOP track to the BEST track, looking for instances of
rapid
> > > > > > intensification. If that's not the case, please let me
know.
> > > > > >
> > > > > > But looking at the DTOP track data, I'm concerned about
the 4th
> > > column
> > > > > with
> > > > > > RI in it, and columns 9 and 10 don't have wind speeds and
> pressure
> > > > values
> > > > > > in the range I'd expect. My guess is that these RI lines
contain
> > > > 24-hour
> > > > > > differences, but the logic in MET hasn't been set up to
process
> > this
> > > > type
> > > > > > of data.
> > > > > >
> > > > > > Instead, MET expects to process raw ATCF track data. Let
me work
> > > > through
> > > > > > an example...
> > > > > >
> > > > > > (1) Use the BEST track for Maria that you sent.
> > > > > >
> > > > > > grep BEST test.dat > bal152017.dat
> > > > > >
> > > > > > (2) Pull the ADECK tracks for Maria:
> > > > > >
> > > > > > wget
http://ftp.nhc.noaa.gov/atcf/aid_public/aal152017.dat.gz
> > > > > >
> > > > > > gunzip aal152017.dat.gz
> > > > > > (3) Run the MET tc_pairs tool to match up the
forecast/best
> tracks:
> > > > > >
> > > > > > met-6.0/bin/tc_pairs -adeck aal152017.dat -bdeck
bal152017.dat
> > > > -config
> > > > > > TCPairsConfig -out al152017_tcpairs
> > > > > >
> > > > > > Running using the default TCPairs configuration file
results in
> > 2,306
> > > > > > forecast tracks being processed.
> > > > > >
> > > > > > (4) Run the MET tc_stat tool to look for RI events but
only for
> the
> > > > OFCL
> > > > > > forecast:
> > > > > >
> > > > > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job
rirw
> -amodel
> > > > OFCL
> > > > > > -out_line_type CTC
> > > > > >
> > > > > > Here's the 2x2 contingency table counts:
> > > > > >
> > > > > > TOTAL FY_OY FY_ON FN_OY FN_ON: 60 1 0 21 38
> > > > > >
> > > > > > Out of 60 points, OFCL got the RI correct once, it missed
it 21
> > > times,
> > > > > and
> > > > > > 38 times no RI was forecast or occurring.
> > > > > >
> > > > > > (5) Now rerun the same job for ALL models present using
the "-by
> > > > AMODEL"
> > > > > > option:
> > > > > >
> > > > > > met-6.0/bin/tc_stat -lookin al152017_tcpairs.tcst -job
rirw -by
> > > AMODEL
> > > > > > -out_line_type CTC -out al152017_rirw_ctc.out
> > > > > >
> > > > > > And write the results to an output file. You'll see that
we now
> > have
> > > > RI
> > > > > > contingency table counts computed separately for each
mode.
> > > > > >
> > > > > > There's many more options here you could play with:
> > > > > >
> > > > > > - Set -out_line_type CTC,CTS,MPR to dump out the
contingency
> table
> > > > counts
> > > > > > (CTC), statistics (CTS), and very detailed information for
each
> > time
> > > > > point.
> > > > > >
> > > > > > - Specify more columns for stratifying the data... add "-
by LEAD"
> > to
> > > > > break
> > > > > > results down by lead time.
> > > > > >
> > > > > > - Change the definition of RI that you'd like to apply.
Perhaps,
> > > OFCL
> > > > > > would score better when defining RI as a 25kt increase
instead of
> > 30?
> > > > > >
> > > > > > Hope this example helps get you going using the MET-TC
tools to
> > > > quantify
> > > > > > RI.
> > > > > >
> > > > > > Please let us know what other questions you have.
> > > > > >
> > > > > > Thanks,
> > > > > > John Halley Gotway
> > > > > > On Thu, Sep 21, 2017 at 8:50 AM, Matthew Onderlinde - NOAA
> Federal
> > > via
> > > > > RT <
> > > > > > met_help at ucar.edu> wrote:
> > > > > >
> > > > > > >
> > > > > > > Thu Sep 21 08:50:25 2017: Request 82076 was acted upon.
> > > > > > > Transaction: Ticket created by
matthew.onderlinde at noaa.gov
> > > > > > > Queue: met_help
> > > > > > > Subject: ERROR : AsciiHeader::read() -> trouble
reading
> file
> > > > > > > Owner: Nobody
> > > > > > > Requestors: matthew.onderlinde at noaa.gov
> > > > > > > Status: new
> > > > > > > Ticket <URL: https://rt.rap.ucar.edu/rt/
> > > > Ticket/Display.html?id=82076
> > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Hello,
> > > > > > >
> > > > > > > I'm trying to use MET-TC to do some rapid
intensification
> > > > verification.
> > > > > > > Here is the command I'm issuing:
> > > > > > >
> > > > > > > $> tc_stat -lookin test.dat -config TCStat_DTOPS_Config
> > > > > > >
> > > > > > > which returns:
> > > > > > >
> > > > > > > DEBUG 1: [Source 1] Lookin: test.dat
> > > > > > > DEBUG 1: Config File Default:
> > > > > > > /usr/local/share/met/config/TCStatConfig_default
> > > > > > > DEBUG 1: Config File User: TCStat_DTOPS_Config
> > > > > > > DEBUG 2:
> > > > > > > DEBUG 2: Processing Job 1: -job rirw -amodel DTOP
-amodel DTOP
> > > > -bmodel
> > > > > > BEST
> > > > > > > -storm_id AL152017 -basin AL -cyclone 15 -rirw_time
240000
> > > > -rirw_exact
> > > > > > true
> > > > > > > -rirw_thresh >=30.0 -rirw_window 000000 000000
-out_line_type
> CTC
> > > > > > > -out_line_type CTS -out_alpha 0.05000
> > > > > > > ERROR :
> > > > > > > ERROR : AsciiHeader::read() -> trouble reading file:
> > > > > > > ERROR : /usr/local/share/met/table_fil
> > > es/met_header_columns_AL,.txt
> > > > > > > ERROR :
> > > > > > >
> > > > > > > My config file (TCStat_DTOPS_Config) and my data file
> (test.dat)
> > > are
> > > > > > > attached.
> > > > > > >
> > > > > > > It seems MET is not properly identifying the header file
to
> use.
> > > > > Could I
> > > > > > > get some help troubleshooting this issue?
> > > > > > >
> > > > > > > To clarify, on my system, the path
> /usr/local/share/met/table_fil
> > > es/
> > > > > > > contains these met_header_columns_VX.X.txt files:
> > > > > > >
> > > > > > > met_header_columns_V2.0.txt
met_header_columns_V3.0.txt
> > > > > > > met_header_columns_V4.0.txt met_header_columns_V4.2.txt
> > > > > > > met_header_columns_V5.1.txt met_header_columns_V6.0.txt
> > > > > > > met_header_columns_V3.0.1.txt
met_header_columns_V3.1.txt
> > > > > > > met_header_columns_V4.1.txt met_header_columns_V5.0.txt
> > > > > > > met_header_columns_V5.2.txt
> > > > > > >
> > > > > > > Thanks,
> > > > > > >
> > > > > > > Matt
> > > > > > >
> > > > > > >
> > > > > > > *Matthew Onderlinde*Meteorologist/Programmer
> > > > > > > NHC/Technology & Science Branch
> > > > > > > (o) 305.229.4466 <(305)%20229-4466> <(305)%20229-4466>
> > > > > > > <http://innovim.com/wp-content/uploads/BD-F-002a-Email-
> > > > > > > Signature-Block-for-Gmail.html#>
> > > > > > > Matthew.Onderlinde at noaa.gov
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
------------------------------------------------
More information about the Met_help
mailing list