Skip to main content

SeqPlot suggestions

We recently released a significant update to the backend systems for the AAVSO website. While most of the bugs introduced by this update have been fixed, there may still be problems we haven't fixed. If you run into a problem, please email webmaster@aavso.org
mrv's picture
mrv
Offline
Joined: 2010-12-21

I have a few suggestions for improvements to SeqPlot.

1) The text within the program is rather small and hard to read on my 1920x1200 screen.  I have the DPI settings on my computer set to 200% of normal.  This works with most of my programs to increase text size but not with SeqPlot.  The info box at the bottom of the SeqPlot window is especially hard to read.  Is there a way for SeqPlot to use the Windows display settings to adjust the text size in the program?

2) The info box shows the number of observations for each star and the source of the data.  The text file only shows the source and not the number of observations.  Could this info be added to the text file as well?

3) After opening SeqPlot and entering the star name and other parameters, I click on the Get Plot button.  There is usually a delay of 5 to 30 seconds while SeqPlot fetches the comp star data.  During this time, the CPU usage on my computer is at 100%.  Why is this?  I'm using a Windows XP laptop with a fairly recent version of Java.


Thanks,

Bob

SeqPlot Suggestions
BSJ's picture
BSJ
Offline
Joined: 2010-03-12

Hi Bob,

Thank you for the suggestions regarding SeqPlot. I will look into what it will take to implement them as soon as I can. Hopefully, it will be easy. Unfortunately, when it was first developed, SeqPlot was designed for internal AAVSO HQ or sequence team use only, so user-friendliness did not figure very high in the requirements. Obviously, this has changed and I need to focus more on how to make it easier to use.

I'm not sure about your computer's CPU usage or how/if it is possible to reduce that, but when you click the "Get Plot" button, what happens is that your computer has to connect to a mysql table in our cloud server, search through about 56 million observations in the database containing the photometry for the relavant data, then plot the results. It can be slower or faster depending on how many sources you select, the field size and the limiting magnitude.

-Sara

Seqplot Changes - User Friendly concerns
CTX's picture
CTX
Offline
Joined: 2010-07-08

My remarks originate with my experience as a member of both the sequence team & the mentoring team.

One of the major potential problems with observers using the Seqplot tool directly for their own sequence is that they typically are not aware of the calibration limits of specific instruments/surveys; that is minimum useable magnitude and maximum useable magnitude; which, can vary, depending upon seeing, with some of the instruments which then burdens the observer with then relying upon a combination of uncertainty and number of observations to judge whether or not a comp is useable. Newer users, from my experience, tend to ignore this consideration and simply select available comps.

Observers new to seqplot and even some of those not so new, do not realize that it is not a planetarium program; the visual position of the stars to one another are approximations and not all the stars in a given fov are necessarily shown.

The failure to understand this last constraint leads observers to often select comps, for their own indivudal use,  that are double and triple systems since they fail to check with any of the various Digital sky image options available elsewhere in order to determine if close in stars accompany a given comp shown in SeqPlot.

Another issue, given that many of the FOV’s have multiple survey data available, would then be the random mixing of that data when an observer creates their own sequence; this may or may not lead to data inaccuracy, especially, if the observer is inconsistent using the data on subsequent submissions.  In an ideal world the surveys would agree within a few thousands of one another;  however, reality is not always in conformity with the ideal concept.

Obviously, a mixing of survey data is very much justified when needed to extend the sequence fainter or brighter than the selected survey might allow (sometimes this can be tricky on the fainter end as the question might arise as to why the survey that goes fainter was not more extensively used as the original selection?).

Therefore, if attempts are made to make this tool more user friendly for observers it would be necessary to create an instruction set, IMO,  listing the limitations/conditions that an observer needs to know in order to use this powerful tool properly; preferably with some examples. 

I am a great believer that an observer should have access to all the tools they need to accomplish and understand their observations; however, in this case, I feel compelled to raise a question as to why an observer would need to directly use Seqplot to accomplish their photometry?  I immediately agree that a nova or super nova could create an immediate demand for a sequence by an indivudual observer.

We have two existing tools now for observers to request a sequence where none is available or to extend or expand an existing sequence by knowledgeable users of SeqPlot:

http://www.aavso.org/chet

http://www.aavso.org/request-comparison-stars-variable-star-charts

In addition, I would question whether or not there Is there is really sufficient justifiable demand by individual observers to justify staff’s time being invested?

I would caution the reader not to presume that I am in opposition to making the tool more user friendly, as I am not.  I am simply raising issues that need to be considered before going down that path, if a decision is made to do so. 

Tim Crawford, CTX

Re: SeqPlot Suggestions
mrv's picture
mrv
Offline
Joined: 2010-12-21

Sara,

Thanks for the reply.  I think most of the interface for SeqPlot is fine.  It was just that some the text was rather small (the star info box at the bottom) and that the number of obs. for each star should be included in the text file.

Regarding the high CPU usage, my impression is that it was a small bug in Seqplot or some odd interaction with JAVA.  It seems to me that the CPU should be idling while its waiting for the comp star data.


Tim,

The reason I use Seqplot is that it contains newer field photometry from BSM, APASS, etc. that wasn't used when the chart team created most of the sequences.  For example, many of the comp star mags in the "official" sequence for SS Cygni use photometry from Tycho-2 and TASS.  I always check to see if newer field photometry is available for a variable before I submit new observations.

Bob

Substitution of Newer Surveys Using Seqplot-Potential Falacy
CTX's picture
CTX
Offline
Joined: 2010-07-08

Bob,

You are an advanced observer and I know you know what you are doing and therefore my remarks are really intended for others as a result of one of your statements that could potentially mislead the casual observer.

I always check to see if newer field photometry is available for a variable before I submit new observations.

Newer potential sequence data does not necessarily mean better data than that shown by using the VSP and it is somewhat of a fallacy to presume so. The majority of the surveys, past or present, are the result, root wise, pretty much, of similar all sky calibration methods & techniques (yep, there are some exceptions). 

Much of the BVRI data, or portions thereof, from many of the surveys, are based upon conversion formulas, some of them being possibly better conversions than others for some wavelengths (FYI, the Rc & Ic data from APASS is based upon conversion formulas from Sloan filters).  All surveys have limits to the useable data on either end of the magnitude scale; with the seqplot tool, the limitations are not always so obvious.  APASS data while quite good does contain a small percentage of FOV’s where the data, for whatever reason, is less than ideal and this might not be so obvious to the casual observer.

Each field calibration (sequence source) has to stand on it's own merits, within known constraints, and the age of the calibration should not necessarily be considered to be a relevant factor.

IMO, the gold standard for sequence data is Henden USNO 1meter (aka #10 superscript with sequence data)

I think it would be a mistake to replace this data with any other survey, especially one that might be considered to be newer; unless a very compelling reason exists (not sure what it might be at the moment).

APASS is not the only AAVSO scope that has been used for calibrations, either; the Sonoita Research scope (aka #18) has been used for quite a few number of sequences and, unless obvious errors are present, there is no real reason to substitute APASS or BSM data (unless extending the sequence brighter); other AAVSO scopes have also been used for calibration work, some of that data could very well be newer than some of the original APASS data.

The most obvious sequence data (brighter magnitudes) that should be replaced, especially for CCD photometry, in all most all cases is TASS.  Either BSM data is fine or Tycho, depending upon which is available [BSM does go fainter than Tycho (which should be avoided, if possible, when the data is fainter than about 10.4)]

Beyond this caution is strongly advised. There will be exceptions from time to time and I acknowledge this but again, I want to caution any user of seqplot data to be knowledgeable about that data and not to simply substitute VSP data because they believe one survey is newer than the one being used for VSP display.

As a general consideration, observers using sequence data other than that available with the VSP run the risk of having their data out of step with other observers; while this could create some confusion with some of the data users it is probably not a major problem in the final analysis, provided the choices the observer makes in their methodology are consistent from observation to observation.

It is the changes that we measure from observation to observation which contain the scientifically useful data; therefore consistency of measurement techniques is probably more important than the specific approach used by any one individual observer (or at a minimum, equally important).

Tim Crawford, CTX

PS – sorry it took me a few days to get back to this thread… just busy.

AAVSO 49 Bay State Rd. Cambridge, MA 02138 aavso@aavso.org 617-354-0484