[teqc] How to use "-O.e[nd] Y M D h m s"

[Lou Estey]

Dear Swen,

> I would like to add "TIME OF LAST OBS" to a RINEX header
> by using the option "-O.e[nd] Y M D h m s".
> 
> All the following commands are without any success:
> 
> a) /usr/local/bin/teqc-32bit-static-2009Mar23 -O.e[nd] 2009 6 18 6 50 30 MARN1690.09O
> b) /usr/local/bin/teqc-32bit-static-2009Mar23 -O.e     2009 6 18 6 50 30 MARN1690.09O
> c) /usr/local/bin/teqc-32bit-static-2009Mar23 -O.end   2009 6 18 6 50 30 MARN1690.09O

The idea is that any of these should work.  The '[...]' stuff (on any option)
is not required and is put into the help output only for readability of the
different options.  In fact, if you try using -O.e during translation, it
works just fine.

>   a) Causes only the message: /usr/local/bin/teqc-32bit-static-2009Mar23: No match.

This message is not from teqc, so it must be from whatever shell you are
using.  (Maybe the '[...]' in the command line is messing up what the shell
thinks is the actual command.  I've never seen this before and can't
duplicate it.)

> b+c) There is no warning, no error and no output of the "TIME OF LAST OBS".
> 
> Where is my mistake?

No mistake on your part.  This is definitely not working for a RINEX-to-RINEX
filtering (i.e. inputting RINEX and outputting an "edited" version of that RINEX).
I'll look into it.

> PS: Is there any possibility to count the number of epochs?

qc the file, e.g.
[3174] teqc +qcq MARN1690.09O 2> /dev/null | grep -i epochs
Poss. # of obs epochs   :      2
Epochs w/ observations  :      2
Epochs repeated         :      0  (0.00%)
Obs w/ SV duplication   :      0  (within non-repeated epochs)
epochs w/ msec clk slip : 0

> What I would like to check is:
> number observed epochs == theoretical number of epochs?
> using the following way:
> -------------------------------------------------------
> + TIME OF FIRST OBS ......... ok: content of each rinex header
> + TIME OF LAST OBS .......... my problem from above
>> compute the difference 
> 
> + INTERVAL (e.g. 30sec) ..... ok: no problem to find out &
>                                   no problem to write into the header
>> compute the theoretical number of epochs
> 
> + the number of epochs ...... ??: Is there any way to count it by using teqc?

That's what the qc is for:
http://facility.unavco.org/software/teqc/tutorial.html#sec_21
but for a complete picture you'll need the corresponding RINEX nav
file(s).

Now in the next teqc release there will be a new line output,
"possible missing epochs", when using the '+meta' option on raw
data (and maybe certain constructs of RINEX obs files), e.g. this
is for a 1-hr file that our polar group is testing with an
auto-reboot on the receiver every hour (causing it to miss a
couple of epochs):

[4228] teqc -notice +meta *.tgd
filename:                test200906051100a.tgd
file format:             Trimble .dat
file size (bytes):       82082
start date & time:       2009-06-05 11:00:00.000
final date & time:       2009-06-05 11:59:30.000
sample interval:         30.0000
possible missing epochs: 2
4-char station code:
station name:
station ID number:
antenna ID number:       00000001
antenna type:
antenna latitude (deg):      40.061198
antenna longitude (deg):   -105.205560
antenna elevation (m):       1572.200
antenna height (m):      0.0000
receiver ID number:      4503244054
receiver type:           TRIMBLE NETRS
receiver firmware:       0.3-1

I might end up disabling this when doing '+meta' on RINEX as input,
though.  The reason is that teqc does not typically read the entire
RINEX file to extract metadata: it reads the header, the first epoch,
and then (if reading a file, and not stdin) usually skips to the end
of the file to find the last epoch.  Doing the qc is still the
best approach.

>>> final check: number observed epochs == theoretical number of epochs
>    to decide if the file is complete or not.

It's all in the qc results.

cheers,
--lou