Looking at data from the archive file

The WXM data is stored in a file called the "archive file". This file is stored in IPW format (the IPP over the wire format, which is architecture independent).

There is a program called read_archive, which allows a user to extract and examine the information from that file. The syntax for calling read_archive is not really fixed (the program is a collection of hacks and outdated ideas). But it is serviceable, and allows plots to be made using the "gnudl mode" option -g.

• read_archive synopsis
• How to read what you want
• Extracting photon tags

read_archive synopsis

Directly from the source code:

usage: gnudl [ -? ] [ -q | -g | -x | -f ] [-l lower] [-u higher] filename
         -q quick mode: shows headers and one-line info
         -x begins to extract information
         -f displays full information (histograms, ...)
         -g display information so GNUDL can use it
         -l lower, -u upper: specify search bounds
         -? displays this information message
     NOTE: if you do not specify -h or -x, it assumes -h all
     NOTE: if you give "-" as a filename, read_archive will read
           standard input, beginning on an IPP message boundary.

How to read what you want

Here are examples of how I use the read_archive program to see what has been happening.

• Monitor what is happening as data comes down. This can be done with

  tail -f archive_whatever | read_archive -x -- -
  

  and you will then see data as it is coming.
• Extract the time history "counts versus time" information, and so that I can then pipe those two columns into another program:

  read_archive -g archive_823202577 | fgrep TH | awk '{print $3 "    " $4}'
  

• Plot the time histories. This can be done with gnuplot or gnudl (using the hete.scm package for gnudl); here's an example with gnuplot:

  <91 sst10b->960201> gnuplot
  
          G N U P L O T
          unix version 3.5 (pre 3.6)
          patchlevel beta 162
          last modified Sun Mar  5 12:10:10 MET 1995
  
          Copyright(C) 1986 - 1994
          Thomas Williams, Colin Kelley and many others
  
          Send comments and requests for help to info-gnuplot@dartmouth.edu
          Send bugs, suggestions and mods to bug-gnuplot@dartmouth.edu
  
  Terminal type set to 'x11'
  gnuplot> plot '<read_archive -g ar* | fgrep TH' using 3:4 with lines  
  

  Here's an example with gnudl:

  Not here yet ...
  

• Look at the commands sent to the WXM and at what time they were sent:

  <113 sst10c->960201> read_archive -g archive_823202577  | fgrep WXM_CMD | fgrep sent
    WXM_CMD 01:49:05.402 6545.402  sent by 7005  is  30  --  0x1e
    WXM_CMD 01:49:05.515 6545.515  sent by 7005  is  31  --  0x1f
    WXM_CMD 01:49:06.649 6546.649  sent by 7005  is  32  --  0x20
    WXM_CMD 01:49:07.573 6547.573  sent by 7005  is  33  --  0x21
    WXM_CMD 01:52:44.315 6764.315  sent by 7005  is  33  --  0x21
    WXM_CMD 02:03:39.906 7419.906  sent by 7005  is  34  --  0x22
    WXM_CMD 02:20:25.485 8425.485  sent by 7005  is  35  --  0x23
  <114 sst10c->960201> 
  

  The commands sent in this example are kind of bogus (30, 31, ...), but here's how to read this line:
  • The first column is there so it can be extracted with grep WXM_CMD.
  • The second column contains the time since last spacecraft boot, in human readable form.
  • The third column contains the same time as the second column, but expressed as a scalar in seconds. This is useful for plotting programs.
  • The number 7005 (in the sixth column) is the process (in this case "tickle") which told XDSP to send the command to WXM.
  • The number in the eighth column (30, 31, ...) is a decimal representation of the command sent. The next number is the same number in hex.

Extracting photon tags

There is a separate program `tag_extract' which extracts the photon tag data from the archive. The session will look like:

tag_extract < archive_file
(00:41:38.226 098 03 11)
(00:41:38.349 006 00 05)
(00:41:38.390 052 09 11)
(00:41:38.431 078 18 05)
(00:41:38.513 072 22 02)
(00:41:38.554 082 18 08)
(00:41:38.636 095 00 05)
(00:41:38.677 032 07 04)
(00:41:38.718 025 04 08)
(00:41:38.759 089 20 05)
(00:41:38.800 010 00 05)
(00:41:38.841 079 26 08)
(00:41:38.882 002 17 04)

The first field is spacecraft time, in hours, minutes and seconds.

The second field is the position (uncalibrated), normalized to the range 0 to 127.

The third field is the energy bin, which ranges from 0 to 31.

The fourth field is the wire, 0 to 11.

You can then look at the time histories (or light curves) from the tag data stream with the command

tag_extract < archive_file | tag2th [-d delta_t]

Here's an example of its use with a 4.5 second delta_t (time resolution):

<10 sst10b->19960809> tag_extract < ../19960807/archive_839443701 | tag2th -d 4.5
 2501.379 000158
 2505.879 000148
 2510.379 000164
 2514.879 000156
 2519.379 000170
 2523.879 000166
 2528.379 000170
 2532.879 000150
 2537.379 000164
 2541.879 000164
 2546.379 000174
 2550.879 000154
 2555.379 000172
 2559.879 000164
 2564.379 000152
 2568.879 000160
 2573.379 000156

The first column is the time, the second column is the total number of counts in that time bin.