[daip] Split CSLGET error
Arnaud Collioud
Arnaud.Collioud at obs.u-bordeaux1.fr
Mon Jun 17 09:49:53 EDT 2013
Le 7 juin 2013 à 16:35, Eric Greisen a écrit :
>> Thanks for the explanation. I now understand the reason why I probably have less split files.
>> So I must have CL entries for each source (if possible) in order to have a split file. I must carefully check my BS table since the corrections I use come from MK4IN.
>
> You could try SNPLT specifying the sources that do not produce output. PRTAB also has ways to print only rows matching some values. Column 3 is the source number so RPARM = 3 , 1, <N>, 0.1 will print only rows for source N and you can find the source numbers with a PRTAB of the source table.
>
>>> I have another user who is claiming difficulties different from yours but indicating that there may still be problems with this code. Your difficulties however suggest that you do not have the latest version.
>
> The issue of that user was a special one. He had entries in the CL table every second and the latest version of the software used 2 seconds as a suitable amount by which to extend times. I have now changed that to 0.4 seconds available today with the MNJ.
>
> Does SPLIT still go off the end of the CL table?
>
> Eric Greisen
>
Dear Eric,
Sorry for the late answer but I decided to start again my processing from scratch, and carefully checking all the steps. Here are some remarks...
/1/
For my session, the tsys measurements are not included in the raw data (because the data is correlated at Haystack, then imported in AIPS thanks to MK4IN), so I am using ANTAB to insert them.
The problem is that, for some stations, the tsys measurements are not well sampled in time. For example, if there is no tsys measurement during a given scan (as defined in the NX table), there will be no correction in the SN table after the run of APCAL.
So I re-sampled all the tsys I have on a regular basis (30 seconds, as the CL table interval). The problem now is that I have too much tsys for AIPS (more than 30 stations times 1500 or so tsys).
Is it possible that you increase the value of the parameter 'MXKINX' in the 'TYCONS' function (file 31DEC13/APL/PGM/NOTST/ANTAB.FOR) from 40000 to 100000 or 200000 for example?
/2/
I still have the problem of CSLGET, but only for single source:
1 2 17-JUN-2013 12:14:56 SPLIT Create 1706-174 .SPLIT . 6 (UV) on disk 1 cno 535
1 6 17-JUN-2013 12:14:57 SPLIT SPLCOP FOUND NO SAMPLES OUT OF 2317 POSSIBLE
1 6 17-JUN-2013 12:14:57 SPLIT IS SUBARRAY 1 IN CL/SN TABLE?
1 8 17-JUN-2013 12:14:57 SPLIT TABIO: BAD LRNO= 50581 LIMIT 50580
1 8 17-JUN-2013 12:14:57 SPLIT CSLGET: ERROR 2 READING CALIBRATION TABLE
1 4 17-JUN-2013 12:14:57 SPLIT Destroyed 1 extension files of type NX
1 4 17-JUN-2013 12:14:57 SPLIT Destroyed UV image file: catno= 535 disk= 1
1 6 17-JUN-2013 12:14:57 SPLIT PROBLEM WITH SOURCE: 1706-174
Please notice that the LIMIT number is the number of lines in the applied CL table (as shown by PRTAB).
For this source, I have 2 majors scans, but the second one is very fragmented, with parts which last 1 second or less (here is the LISTR output):
372 1706-174 : 0000 2 1/04:29:13 - 1/04:37:29 1 2381213 2429194 (Scan 1)
891 1706-174 : 0000 2 1/17:28:43 - 1/17:28:44 1 4858477 4858538
893 1706-174 : 0000 4 1/17:28:22 - 1/17:29:27 1 4857422 4860742
894 1706-174 : 0000 2 1/17:29:05 - 1/17:29:06 1 4859599 4859660
896 1706-174 : 0000 2 1/17:29:27 - 1/17:29:28 1 4860716 4860776
898 1706-174 : 0000 2 1/17:29:49 - 1/17:29:50 1 4861827 4861887
900 1706-174 : 0000 3 1/17:29:28 - 1/17:30:45 1 4860782 4864517 (Scan 2)
901 1706-174 : 0000 2 1/17:30:11 - 1/17:30:12 1 4862938 4862998
903 1706-174 : 0000 1 1/17:30:45 - 1/17:30:45 1 4864484 4864524
904 1706-174 : 0000 2 1/17:30:45 - 1/17:30:45 1 4864491 4864547
907 1706-174 : 0000 2 1/17:31:07 - 1/17:31:07 1 4865119 4865153
909 1706-174 : 0000 2 1/17:31:29 - 1/17:31:29 1 4865724 4865758
911 1706-174 : 0000 1 1/17:30:46 - 1/17:32:47 1 4864549 4866824
912 1706-174 : 0000 2 1/17:31:51 - 1/17:31:51 1 4866272 4866299
I do not understand because I have corrections for this source, at least for the first scan, as displayed below (extract from the applied CL table):
COL. NO. 1 2 3 4 5 6 7 8
ROW TIME TIME INTER SOURCE ID ANTENNA NO SUBARRAY FREQ ID I.FAR.ROT GEODELAY
F NUMBER D/HMS D/HMS RAD/M**2 SECONDS
21365 04:29:13.0 00:00:30.0 197 2 2 1 0.000000E+00 0.0000000000D+00
21366 04:29:13.0 00:00:30.0 197 10 2 1 0.000000E+00 0.0000000000D+00
21367 04:29:13.0 00:00:30.0 197 11 2 1 0.000000E+00 0.0000000000D+00
21368 04:29:13.0 00:00:30.0 197 12 2 1 0.000000E+00 0.0000000000D+00
21369 04:29:13.0 00:00:30.0 197 17 2 1 0.000000E+00 0.0000000000D+00
21370 04:29:13.0 00:00:30.0 197 27 2 1 0.000000E+00 0.0000000000D+00
21371 04:29:13.0 00:00:30.0 197 30 2 1 0.000000E+00 0.0000000000D+00
21372 04:29:13.0 00:00:30.0 197 31 2 1 0.000000E+00 0.0000000000D+00
21393 04:29:43.0 00:00:30.0 197 2 2 1 0.000000E+00 0.0000000000D+00
21394 04:29:43.0 00:00:30.0 197 10 2 1 0.000000E+00 0.0000000000D+00
21395 04:29:43.0 00:00:30.0 197 11 2 1 0.000000E+00 0.0000000000D+00
21396 04:29:43.0 00:00:30.0 197 12 2 1 0.000000E+00 0.0000000000D+00
21397 04:29:43.0 00:00:30.0 197 17 2 1 0.000000E+00 0.0000000000D+00
.....etc.
21792 04:37:13.0 00:00:30.0 197 11 2 1 0.000000E+00 0.0000000000D+00
21793 04:37:13.0 00:00:30.0 197 12 2 1 0.000000E+00 0.0000000000D+00
21794 04:37:13.0 00:00:30.0 197 17 2 1 0.000000E+00 0.0000000000D+00
21795 04:37:13.0 00:00:30.0 197 27 2 1 0.000000E+00 0.0000000000D+00
21796 04:37:13.0 00:00:30.0 197 30 2 1 0.000000E+00 0.0000000000D+00
21797 04:37:13.0 00:00:30.0 197 31 2 1 0.000000E+00 0.0000000000D+00
21800 04:37:29.0 00:00:30.0 197 2 2 1 0.000000E+00 0.0000000000D+00
21801 04:37:29.0 00:00:30.0 197 10 2 1 0.000000E+00 0.0000000000D+00
21802 04:37:29.0 00:00:30.0 197 11 2 1 0.000000E+00 0.0000000000D+00
21803 04:37:29.0 00:00:30.0 197 12 2 1 0.000000E+00 0.0000000000D+00
21804 04:37:29.0 00:00:30.0 197 17 2 1 0.000000E+00 0.0000000000D+00
21805 04:37:29.0 00:00:30.0 197 27 2 1 0.000000E+00 0.0000000000D+00
21806 04:37:29.0 00:00:30.0 197 30 2 1 0.000000E+00 0.0000000000D+00
21807 04:37:29.0 00:00:30.0 197 31 2 1 0.000000E+00 0.0000000000D+00
Any idea why I have no output for this source?
(I also tried without flagging, but that changed nothing)
/3/
> Does SPLIT still go off the end of the CL table?
Yes, it seems so since I have data for the last source of the session.
Best regards,
Arnaud
More information about the Daip
mailing list