Affiliation
American Association of Variable Star Observers (AAVSO)
Wed, 05/13/2015 - 01:30
What does Boxter expect as an input?
If I use TA to generate a transformed but not aggregated report that includes six observations, three in V and three in I, save that report as a text file, and try to open that file using Boxter, I get an error box that says “Index was outside the bounds of the array.” What does that mean?
I believe TA is needed to first transform the observations but I can't figure out what TA output to feed Boxter.
Unlike TA, it appears Boxter wants to open a file via the File menu instead of accepting a paste.
Is there a description of the workflow or a cookbook for taking observations from VPHOT, through TA, through Boxter, and to WebObs for submittal?
Thanks,
Jim Seargeant
Hello Jim,
"Boxster" is a utility program that opens AAVSO Extended File format text files containing time series observations of a single star through a single filter band. It allows you to combine an arbitrary number of contiguous observations into a shorter list of observations.
A companion program zipped into the same file on the AAVSO website is named AAVSO Utility (sorry for the meaningless name). It opens an AAVSO Extended file report containing more than one star and more than one filter band for each star. The program combines all observations of each Star-Filter combination and calculates the appropriate mean magnitude and error (Std Error of the Mean) for each group. It outputs a new Extended file format text file that you can upload to WebOps.
I hope this helps.
Richard Sabo
TA does offer aggregation of observations. It is simpler than the Boxter/AAVSO Utility tools in that always aggregates all the records its sees in the input if they have the same star name, filter, comp and check star.
Note also that this aggregation happens before the transforms are applied. I'm not sure if there is a significant advantage to doing the aggregation post transformation. I can add that option if there is strong interest.
George
Thank you Richard and George.
What I'm trying to do is to build a workflow that will move data from observations to a WebObs submittal that will satisfy Arne's criteria for acceptable observations for his Anomalous Cepheid Campaign. I'm new at this, so it takes some study for me to figure out how to string the tools together.
From Arne....
“Please transform your data, taking at minimum either B&V or V&Ic with every dataset.”
“I am going to request that you take a minimum of 3 sets of data, average them and report the mean and standard deviation. A program like Boxster can do this for you.”
For this exercise, I've taken the three sets of data in V and Ic and transformed them with TA.
I tried to aggregate them to find the mean and standard deviation via Boxter because that's what Arne suggested.
AAVSO Utility …..
Richard, based on your response, Boxter is the wrong tool to use on my “nightly snapshot” of 3ea shots in V and Ic unless I split the V observations from the Ic. I tried AAVSO Utility, but when I opened the TA output report (see BLB_TAreport_noagg.txt, attached) of three observations, one star, two filters I got the error “ Index was outside the bounds of the array.”
TA............
George, I've been able to aggregate them with TA and see that TA reports the mean of the three observations, but I can't see where TA reports the standard deviation. Does it and if not could that be added? TA would be the easiest tool to aggregate the observations since it's already open for transformations.
Thanks,
Jim Sgt
When TA aggregates the raw observations the VERR field is replaced with the Standard Deviation of the Mean. The final error will be a little larger as the transform equation contributes its error to that value.
I note that the VNAME field in your records is "Star 1". If you are using VPhot you can get that renamed to "BL Boo" from within VPhot, saving you a job step later or changing the name.
You mention you got an index error when feeding AAVSO Utility output into TA. Or was that an error thrown by the AAVSO Utility? If you can describe the error and send me the input and output files I can pin that down. You can do that offline by sending the files direct to SGEO@GASilvis.net.
George
Hello Jim,
The AAVSO Utility only accepts text files in the "Extended AAVSO Format", ie, the same format needed to upload files through WebOps. I'm sure there is a way to produce such a file from TA, but George can comment on that.
When Arne says he would like the uploaded data to contain the mean magnitude and the standard deviation, I think he probably would accept "Standard Deviation of the Mean" also called "Standard Error" which is what the AAVSO Utility calculates as MERR.
He contributed to the discussion of error reporting when I wrote the program and he also recommened that error measurement in his "CCD School".
Richard
Hello Jim
I too have been observing this campaign. Unfortunately, the sky has not cooperated and I have too few observations to agregate.
What I planned to do was to use Boxster first, aggrgate by 3, it gives an AAVSO Extended Format which can then be inputed to TA. I have done this a couple of times before. TA writes a file that can be input to Webobs without Modification.
The only snag may be that TA will get confused when there are more than 1 comp with the same designation in the sequence data base. It does not happen often, but there are a few, and this is one of them. I had this issue with the 140. If you use the AUID rather than the 140 or equivalent, everything should work ok.
Gary
Sorry for the delay in picking up this discussion again. There were a couple of beautifully clear nights that had priority and it took a while to sort this out. I hope this isn't too tedious.
George, I see now how to replace “Star 1” with a descriptive name via the load sequence in VPhot (thanks) but am sticking with the original data for the attached files for now.
(My goodness, I just realized I could do the same thing with the names of the Comp and Check stars. Some of us take a little (a lot) longer to figure things out.)
I've been trying a workflow of VPhot > TA > AAVSO Utility (AU). Gary, you planned to use VPhot (I assume VPhot) > Boxster > TA.
Setting aside the question of Boxster vs. AU, does it matter which comes first, Boxster or TA? Both are supposed to output data in the AAVSO Extended Format (AEF). Both are supposed to accept inputs in the AEF.
(Gary, returning to Boxster vs AU – Richard says Boxster opens ….single star through a single filter band.” AAFSO Utility opens … “more than one filter band for each star”. How do you handle your two-filter observations of BL Boo using Boxster?)
So I tried it both ways.
Trying VPhot>AU>TA.......
Using VPhot I generated the file “BLBOO_from_VPhot_VandI_3ea” (attached).
I opened that file using AAVSO_Util which automatically generated the binned result in the lower window, copied the lower window to the clipboard and also saved the result as “BLBOO_from_AAVSOUtil_VandI_3ea” (attached).
I pasted the clipboard into TA, pressed Process and got the error “”Na” is not a valid floating point value.”
I don't know what's going on there.
Trying VPhot>TA>AU (not aggregating in TA) …..
Long story short, this worked, as did VPhot>TA (aggregating in TA).
What caused my earlier problems was trying to use TA's Report to feed into AU. The lower window of TA, not the Report window, has the AEF data that can be saved for submittal or to input into AU. Should have been obvious if only I had been familiar with what the AEF looks like.
So...
I can get my BL Boo observations transformed and combined using either TA alone or with TA > AU.
There is a slight difference in the results.
Transforming and aggregating in TA yields
Filter Mag MERR
V 14.688 0.004
I 14.385 0.014
Transforming in TA and aggregating in AU yields
Filter Mag MERR
V 14.689 0.003
I 14.385 0.013
For the purpose of working with a single target star and two filters, Is there any reason to use the AAVSO Utility when TA will transform and aggregate?
Thanks for your patience
Jim Seargeant
Hello Jim
I use Boxster n times, once for each filter. Then I paste a file together with the n filters, all in one extended aavso format to input to TA.
As for AU, I have never used it. I'll look into that.
BTW: Maxim 6 outputs one file with all filters included. This is very handy. Makes the aggregation much harder, so I rarely aggregate any more. I would rather just expose longer, with TDM its not a problem up to 5 minutes, and longer than that, I throw on the autoguider.
The Maxim 6 output can be opened in TA in all colors very easily, so makes the work flow very nice. Maxim 6 separates the files and organizes them. I have to open only one nightly folder per object in Maxim 6. Very handy. No fiddling with files, no opening by color, by object etc, that one had to do with Photometric images in Maxim 5.
Gary
Gary
Gary I haen't been using Maxim 6 it's full capability. I'll look into using it for photometry.
After, that is, I get through the course on VPhot.
Jim
Jim,
TA is throwing an error when fed AAVSO Utility output because AU is outputing CMAG= na, it's not averaging the CMAG field. That's a problem. Make sure you check this if you are using AU.
If AU was working, then the issue of running AU after TA or just using TA with aggregation is the issue of when to apply transforms, post or pre aggregation. One difficulty with post transform aggregation is that TA has put a lot of information into the comment field to document how the transform was performed. This is going to be either lost or become a mess in the final record.
I'll take a look at adding a post transform aggregation feature to TA that will address this.
George
Thank you, George.
Jim Sgt
Jim,
The most common cause of the AAVSO Utility program outputing CMAG=na is the presence of an ENSEMBLE used in one or more observations in the input file. All of the observations must have a CMAG as a decimal number in order to calculate the mean magnitude.
If that is not the problem please send me the input file and I'll try to identify the problem.
Richard
rsabo333@gmail.com