[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: suggestion/TASS FOR DUMMIES
- To: "TASS" <tass@wwa.com>
- Subject: Re: suggestion/TASS FOR DUMMIES
- From: "Shawn Dvorak" <sdvorak@bright.net>
- Date: Tue, 14 Apr 1998 20:58:06 -0400
- Old-Return-Path: <sdvorak@bright.net>
- Resent-Date: Tue, 14 Apr 1998 20:59:45 -0400
- Resent-From: tass@wwa.com
- Resent-Message-ID: <"yKR5HB.A.y7D.uYAN1"@kani.wwa.com>
- Resent-Sender: tass-request@wwa.com
Thanks for the response. I guess I'm just looking for too much right now;
it's still very much a work in progress. My biggest complaint is that
there isn't much organization to what *is* there - like the messages posted
to this list like "file x is now in /pub on server-y", without any
information about the format of the data or even the complete address of
server-y (which is fine if you're an old-timer but not much help to the
uninitiated). I appreciate your suggestion about just picking a piece and
asking some questions to understand the data and processes. Okay, I'll
start doing something constructive just as soon as I get relocated in our
new house and get everything settled.
Shawn Dvorak
----------
> From: Chris Albertson <chris@topdog.pas1.logicon.com>
> To: Shawn Dvorak <sdvorak@bright.net>; chrisa@wavenet.com
> Subject: Re: suggestion/TASS FOR DUMMIES
> Date: Tuesday, April 14, 1998 4:47 PM
>
>
> Here is an outline. Perhaps you could pretty it up and point out
> any big holes it may have. If you need more specific info just ask.
> If I don't know I can point you so where it may be. The point that
> Tom made is that if you appear to be doing some work people will
> help. I agree getting started is hard. The idea is "start small"
>
> If you are looking for how data is processed end to end you will not find
> it. Reason: It is not yet being done. This could be the reason for your
> dificulty in unscrabling all the unorganized stuff you we on the web
site.
>
> With the above in mind here goes.
>
> 1) There are a few people who have cameras. Not all are identical. Most
> have three CCDs some have one, one I think now has only two. There are
> three kinds of filters V, R and I., You can figure out who has what be
> looking at the data. ZIP files at the FTP site (see below) I don't think
> there is a master list. I would like to compile one.
>
> 2) Each site collects data with one of two control programs
> a) A DOS based one called tm3get11 or,
> b) A Linux based one called ?? it has no name that I know of
> Most everyone uses "a". These programs each output "FITS Files"
> These are images. (there is a "FITS Standrd" and a whole host of
> programs for dealing with FITS format data on the web.
>
> 3) These FITS files are reduced to ASCII text files called star lists.
> There are any number of programs that can do this. IMO the jury is
still
> out on which one is best but for now everyone is using "STAR.EXE" for
> "production" work. "sextractor", IRAF/DAOPHOT, DoPhot and other
> may work better.
>
> 4) Output from STAR is batched into ZIP files and FTP'd to
> ftp.tass-survey.org/pub/incoming where Michael can get it and process
it into
> his database system. Also most people process these STAR output
locally
> with whatever tools they may have on hand. You will have to ask each
> camera owner what if anything they do with thier data. Not all of the
> collected data is sent, as Michael is "backed up". (I would download
> just one small sample)
>
> 5) The "TASS Database" that Micheal runs is Postgresql. To get at the
data
> you can use his WWW interface or if you have Postgres client software
> (runs on any UNIX/Linux system) you can do SQL queries directly and
bypas
> the WWW interface. Postgresql also supporots ODBC and JDBC. ODBC
drivers
> for win95 are available with Postgresql at http://www.posgresql.org I
have
> used
> MS Excel to access his database using MS Query and OBDC.
>
> I would suggest _not_ trying to set up an end-to-end system. Pick a
point.
> learn how that works then maybe expand. For example maybe you are
interrested
> in the process of converting images to lists and want to first run the
> curent version of STAR then see if you can do better. Or perhaps you
take
> a look at the current WWW interface to the Data and figure you can do a
littel
> better. Ask for details that interrest you. Let's go one file type at a
> time. If you want an image file get a few an spend a couple weeks
learning
> how to look at them that move on to something else. (Images are not to
be
> found on the FTP site except if you ask for a few samples)
>
> Once you have a few images. You may ask "How do I look at these?" IF
you
> do I'll tell you all about IRAF. (http://iraf.noao.edu/) Someone else
will
> tell you about some
> Windows95 IP program he likes. You will likely get a bunch of answers.
>
>
> I am working on two chunks of software and know them pretty well as I am
> the one who is doing the work. Other areas I haven't as much as a clue.
>
> The two areas I can talk about in detail if you want are:
>
> 1) Linux based real-time camera driver. This thing runs under Linux
> and twiddles bits inside the TASS CCD hardware and writs out FITS format
> image files. It takes full advantage of it's UNIX enviroment and is
> designed as a set of client/server processes.
>
> 2) The TASS Database SOftware. I am writting software that reads those
> ZIP'd starlist files and loads them into a Postgres Database system.
Along
> the way the software figures out _what_ stars where observed, updates a
> "master catalog" and so on.
>
> Both the above are functioning now but require more work before I will
> asign a version ID or "1.0".
>
> So if TASS is like playing that old game "Zork" I hope the above is like
> a kind of map. Just explore one area at a time. You can see I am asking
> some very basic questions still. like "how to apply a transform to data
> we don't have?"
>
> One whole area that is not yet addressed by TASS is what I call "Data
> exploitation" That is where we look at the data in Michael's databas and
> try to figure out what it all means. I fugure there are some supper nova
> in there waiting to be discovrred. I haven't any idea (yet) how to
> find them but that's the fun part, figuring out how to discover them.
> One you do that finding them will be an automatic process.
> --Chris Albertson
>
> chris@topdog.logicon.com Voice: 626-351-0089 X127
> Logicon RDA, Pasadena California Fax: 626-351-0699