Editing Scanco Administration (section)

====Measure multiple samples in one scan====

This quite often happens when you want to scan 6 rat bones, for example, but using our six-gun sample holder you scan them in one go.  But, you want to analyse them separately, so you have to import the ISQ multiple times as new MEASUREMENTS into each SAMPLE.

There are two options: manual and auto.  <b>Read this entire HowTo before proceeding.  There is significant potential for mistakes!</b>

<pre>
$ @com:import_manual.com

P1: Measurement Number to duplicate
P2: New Sample Number from Database
P3: New Start Letter of ISQ file
P4: Optional: Operator Number
P5: Optional: Controlfile Number
</pre>

So:

<pre>$ @com:import_manual.com 1234 342 D</pre>

will copy the ISQ from measurement 1234, import it into the database to sample 342, give some feedback lines with 'old' and the new samplename, CHECK IT, and at the end it gives the feedback line on the terminal which you take with the mouse and drop to rename the data over to the new location.

<b>IMPORTANT: BE VERY SURE NOT TO MIX UP MEASUREMENT AND SAMPLE NUMBER !!!</b>

This distiction needs to be VERY clear in your mind! Of course, you can duplicate one scan and give it to the same samplenumber again (e.g. to evaluate different parts of the same sample, and each has its database
entry).

Please build up some system so you see immediately that the right meas. goes to the right sample. E.g. all four bones in one scan have the same part in the name, distinct from other 4 bones. If there is any trace of a doubt later on that somebody MAY have messed up the samples in this duplication step, you'll most probably have to start from scratch again...

Of course, I'll have to make you aware that Scanco declines all responsibility - this program is NOT in the normal user-friendly, quite fool-proof Scanco controlbox, but it's up to you to use it carefully, and not on Friday night when you're sleepy and mistakenly mix up sample and meas. numbers...

The new startletter I built in so you always know which one was the 'original' measurement (which has an RSQ), and which are the duplicated ones.

<b>Backup</b>
And the organization of how you know which duplicated scan is which part is
completely up to you - as well as how you want to handle the backup - our
software allows you to back-up each duplicated scan (but this might be an
overkill, since it's the same ISQ always) so you may choose to delete the
duplicated ISQ by hand in the session manager (only keep the 'original' ISQ), and only backup the .gobj, .aim and _seg.aim etc. (but then you will need to know not to get scared if you retrieve a scan, and the ISQ is missing...) (Or you just 'move to trash' the duplicated scans to nicely delete them with the backup program if you want to remove them completely (but you still have the 3D results in the evaluation database))

<b>Autosubmit</b>

<pre>$ import_autosubmit.com</pre>

Please use the manual one for a while. Then, if you feel comfortable, use the autosubmit one (which also does the last command automatically, the renaming of the isq to the new location.) Because it goes into a batch que, you can quickly submit all duplications, and do not have to wait for the whole file-copying like in the manual one.

One thing you should check there:
$ show que sys$fast/full

This needs to be one single queue (/job_limit=1). In a clustered enviroment,
this might be a '/generic=' que that then gives the jobs to multiple queues in the cluster. This is NOT good. If this is the case, please edit the autosubmit script, and replace sys$fast in the following line at the end by ONE of the generic fast queues, e.g. SC4240_FAST or whatever the name is (preferrable the one computer where also the data disk DK0 is sitting:

<pre>
$!
$      SUBMIT/NOPRINT/QUE=SYS$FAST/DELETE/NOTIFY/LOG=SYS$SCRATCH: 'COMMANDFILE'
$!
</pre>

(Reason: if multiple uct_imports would run at exactly totally the very same time, both jobs might get the same new measurement number... bad...)

I also suggest to add the following line to your login.com

<pre>$ dupl*icate :== @com:import_autosubmit.com</pre>

then you can just start it with

<pre>$ dupl 1234 342 D</pre>

And I also recommend to put into login.com:

<pre>
$ check_dup :== search sys$scratch:import*.log """ name:""
</pre>

(yes, with all these double-quotes)

Then use it to check the feedback that the right scans were given to the right samples.


[[Image:Redirect.jpg|15px]] [[#top|''Return to Contents'']] or [[Main_Page|''Return to Main Page'']]