[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Date: very important.
Sebastian,
I speak for myself, not the group.
Your point may be valid, but it is nearly lost in the antagonistic delivery.
You have presented some potentially useful data to help figure out how much
data may be a problem, and then in the next breath attempt to blackmail the TASS
group to do as you see fit.
Your suggestion to remove data for 2004 should be taken into consideration.
Maybe you can help further by looking for more data that is apparently in error.
One star is not exactly compelling evidence. Many stars showing the same offset
would help your cause, and help the TASS group determine how pervasive the
problem may be.
If you are following the thread, you will notice that action is being taken to
determine what the root of the problem is, and how pervasive it is. Being
constructive rather than abusive will help your cause.
This is a volunteer effort, in all aspects. Are you willing to help track down
the problem?
Rob
When grilled further on (Mon, 14 Jun 2004 22:26:24 -0300),
"Sebastian Otero" <varsao@fullzero.com.ar> confessed:
> Dear colleagues,
>
> As most of you know, some of us are making intensive use of the publicly
> available databases to do systematic searches of variables.
> Last week, large discrepances between TASS and NSVS data for an eclipser
> were noted and Patrick Wils said that it was a date error. I sent a
> lightcurve confirming that and other people also noted the same situation in
> other cases.
>
> While other people continue to use words as "alleged" regarding this
> problem, my colleague Pavol Dubovsky sent me some stars that he analysed and
> one of them --using TASS observations besides NSVS-- had a very important
> remark: "period change according to TASS data".
> I inmediately downloaded the data and made lightcurves only to confirm that
> all TASS eclipses are 1 and 2 days off (YES, TWO).
> This is going out of hand. All dates are a chaos recently.
> Please take this example and stop trying to explain the unexplainable:
>
> NSV 5040 is a 3.02405 d. period eclipser. "NSV5040_shame" shows how all TASS
> eclipses fall outside the expected phases. These are the HJD (2450000+)
> dates and the difference with the real world:
>
> 3085.6800 12.591 ?
> 3096.6637 12.710 ?
> 3096.6654 12.714 ?
> 3098.6460 12.754 +1 or +2
> 3098.6477 12.609 +1 or +2
> 3099.6488 12.899 +2
> 3099.6505 12.859 +2
> 3103.6348 13.308 +1
> 3103.6364 13.027 +1
> 3105.6293 12.753 +1
> 3105.6310 12.763 +1
> 3106.6237 13.347 +1
> 3106.6253 13.341 +1
> 3109.6150 13.181 +1
> 3109.6167 13.125 +1
> 3112.6164 13.030 +1
> 3112.6181 12.928 +1
> 3114.6136 12.629 +1
> 3120.5919 12.686 +1
> 3120.5936 12.711 +1
> 3121.5803 12.705 +1
> 3121.5820 12.720 +1
>
> So it's not a 1-day problem. It's not day 3109 alone.
>
> "?" means that the date could be whatever. ((Maybe TASS is an online version
> of the lotto??))
> "+1 or +2" means that the date is wrong but it might be corrected by adding
> 1 or 2
> "+1" means that the dates are corrected by adding 1 day.
> "+2" means that the dates are corrected by adding 2 days.
>
> "NSV5040_realworld" is the lightcurve with these corrections applied. The
> revelation of the 2 points of day 3099 matching exactly in the ascending
> branch by adding 2 is astonishing.
> I don't know if I should make jokes about this or simply cry....
>
> I urge that you inmediately hold back from the web all TASS data at least
> for these year 2004, because it's a total lack of respect to people having
> this data avaliable online. Not all researchers are subscribed to the TASS
> list and know what is happening. But if the data is not taken off as soon as
> possible, I will take care of publishing this real shame in every way I can
> so they become aware.
> Furthermore, I think that if you are responsible enough, you should make the
> web access unavailable until you completely fix this embarrassing problem.
>
> The TASS database is completely useless right now.
>
> Sebastian Otero.
>
--
20:54:30 up 47 days, 3:38, 5 users, load average: 2.08, 2.17, 2.13
Linux 2.6.5-01 #7 SMP Fri Apr 16 22:45:31 MDT 2004
PGP signature