Post-processing an older cruise
-----------------------------------------

Every effort is made for `quick_adcp.py <../../quickadcp_overview.html>`_

 - to be internally consistent
 - to prompt the user if something is missing
 - to have decent defaults for most common circumstances

If you run the command ``quick_adcp.py --help``, the
processing options should look like this:

::


  To see commands for various data types, use --commands as follows.
  --------------------------------------   ------------------------------
  quick_adcp.py --commands  postproc     :   UHDAS post-processing
  quick_adcp.py --commands  uhdaspy      :   UHDAS processing
  quick_adcp.py --commands  ltapy        :   LTA or STA files (averaged) (**)
  quick_adcp.py --commands  enrpy        :   ENR files (beam coords)
  quick_adcp.py --commands  pingdata     :   original pingdata demo
  ======================================   ==============================

(**) LTA and STA processing can now be processed more quickly using
``vmdas_quick_ltaproc.py``.  Try typing ``vmdas_quick_ltaproc.py --help``
for more information



If the command ``quick_adcp.py --help`` produces any references to
matlab processing, then your programs are old and this version of the
documentation will have limited use.  You *should* have your own copy
of the documentation in the directory where you put the CODAS software.
That is the appropriate resource, because it should be consistent
with your own code.  This would be an excellent time to install
a new Virtual Computer so you are running up-to-date code.


**Storing metadata**

When ``quick_adcp.py`` runs, it caches (stores) some command-line
options in a file called ``dbinfo.txt``.  This file is used to look up
values that do not have defaults, which were specified during the
original run, so you do not have to type them again each time.  There
are some cases in which these values might change: ``quick_adcp.py``
expects a specific format and values in the file, so do not edit this
file by hand.



**Example**

Information about the dataset, such as year and ADCP model, are cached
in a file called "dbinfo.txt" during processing.  This file was only
implemented with Python processsing, so if you are post-processing a
UHDAS dataset that was processed at sea by an earlier version of the
software (eg. using Matlab), then there is no ``dbinfo.txt`` file.
If "dbinfo.txt" does not exist, you must supply the information so
quick_adcp.py can generate the file properly.


The best commands to run are to reset the navigation (improving an
older approach to ship speeds) and recalculate the calibrations.  Each
time you run it, let ``quick_adcp.py`` prompt you (complain) as
another variable is discovered missing: go ahead and specify the
missing information as needed.  Once ``quick_adcp.py`` has enough
information, it will write the values out to ``dbinfo.txt`` and
proceed.  You will probably have to provide values for variables such
as ``yearbase``, ``--cruisename``, ``--sonar``, ``--beamangle``.  This
step is a small but unfortunate requirement for compatibility of newer
software with older processing directories.

Run ``quick_adcp.py --steps2rerun navsteps:calib --auto``

At each error, include the information in the quick_adcp.py
command line until it runs.

+------------------------------------+----------------------------------------+
|         ERROR                      |    solution: add to commands           |
+====================================+========================================+
|   ERROR -- must select datatype    |    "--datatype uhdas"                  |
+------------------------------------+----------------------------------------+
|   ERROR -- must set "sonar"        |    "--sonar os38nb"                    |
+------------------------------------+----------------------------------------+
|   ERROR -- must set "beamangle"    |    "--beamangle 30"  (see **NOTE #1**) |
+------------------------------------+----------------------------------------+
|   ERROR -- must set "yearbase"     |    "--yearbase 2010"                   |
+------------------------------------+----------------------------------------+
|   ERROR -- must set "ens_len "     |    "--ens_len 300"   (see **NOTE #2**) |
+------------------------------------+----------------------------------------+
|   ERROR -- must set "cruisename"   |    "--cruisename  km1001c"             |
+------------------------------------+----------------------------------------+

**NOTE #1**: The variable ``ens_len`` (seconds per averaging enseble)
must match the averaging length in seconds that was used when originally
processing the data.


**NOTE #2**: Beam angles are usually (but not always) as follows:

+---------------------------+------------------------+
|   instrument              |  beam angle (degrees)  |
+===========================+========================+
|  os38, os75, os150        |       30               |
+---------------------------+------------------------+
|   wh300, wh600, wh1200    |      20                |
+---------------------------+------------------------+
| bb75, bb150, bb300, bb600 |  usually 30            |
+---------------------------+------------------------+
| nb150, nb300              |  usually 30            |
+---------------------------+------------------------+


Example:


``quick_adcp.py --steps2rerun calib --datatype uhdas --cruisename km1001c --yearbase 2010 --sonar os38nb --beamangle 30 --auto``