head 1.1; access; symbols; locks http:1.1; strict; comment @# @; 1.1 date 2003.07.03.15.13.14; author MarkusDolensky; state Exp; branches; next ; desc @none @ 1.1 log @Test report on AVO Prototype V0.926 (M. Allen) @ text @Subject: Testing AVO v0.926 From: "Mark G.Allen" Date: Sat, 28 Jun 2003 20:34:48 +0200 To: Markus Dolensky , Pierre Fernique , boch , Anita Richards , Martin , mleoni@@eso.org, Francois Bonnarel , Alberto Micol Hi everyone, I've done some first tests using AVO v0.926. It is looking very good! - but ofcourse I've found some things are broken and some bugs. Here are my notes, plus questions and suggestions. I've organised it into Aladin, SED and ACE components. I've put a * on things I think are critical (mostly obvious). I haven't tested the VOPlot part yet. Test system: Linux Redhat 9.0 1. Aladin RGB: - At first I thought this was not working, but it's just that the menu buttons are new, and require clicking on the right side of the button where the triange is. ACS epoch 3 images: * - The display autoscale is being confused by too many 0s in the data. - The epoch 3 ACS image I looked at have large areas with the value 0 (filling in the edges around the non-square fields) The image data values are lower around 0.001 ZOOM: * - The zoom becomes deselected by moving the mouse over the "Z" button without clicking (by design I'm told by Pierre). That's fine, however, afterwards it is not possible to select catalog points in the window. Sometimes zooming again using the Zoom menu (or even selecting a panel in the stack) fixes this problem, but not always. Data Tree: - Can we take out the uneccesary"epoch" level in the WFI,ISAAC and ACIS of the tree? - General comment: The data tree is really great! The way everything, including local data, external data and stuff from the image server can be intergated uniformly in the same place is excellent, and fairly easy to use. I tested adding the full NTT-SOFI data of the CDFS with no problem. Plane Stack: - When a folder is at the top of the stack, images will load into the folder - is this a feature or a bug? Perhaps it should load into the folder when the folder is open, but load onto the top of the stack when the folder is closed (?). History Tree: - I find it easier to use the history Tree than the tree in the server selector window because it is possible to expand the history tree window and see all (or more) of the tree at the same time. I would suggest we change the name to "Data Tree". Loading local data - I loaded a full 4k x 4k ACS image. This was cut down by Aladin, presumably because of memory. - Note: This is fixed by starting up with allocation of more memory: java -Xmx256m -jar AVO.jar 2. SED: Default mode: - The default flux[Jy] vs. frequency[Hz] mode works fine Axis unit conversions: - When the bottom axis is changed to a wavelength, (Angstrom,metre,micron), the axis label should also change to "wavelength" : e.g. wavelength[Angstrom], not frequency[Angstrom]. Having frequency[Angstrom] would make us look pretty silly in some Astronomers eyes, so I think it's important to fix it. * Axis ranges: - It would be desirable avoid 0.0 in the frequency and wavelength axis ranges. Flux density conversions: - There is no conversion of F-nu(Jy etc.) to F-lambda(erg s-1 cm-2 Angstrom-1) This is more difficult ofcourse, but the most common ways of plotting SEDs are flux[Jy] vs. frequency[Hz] flux[erg s-1 cm-2 Angstrom-1] vs. wavelength[Angstrom] (nu*Fnu was also in the original specs...). If there is time to add this capability, remember the IDL SED test tool can be used to check that the conversions are done correctly. (http://www.euro-vo.org/internal/Avo/WorkPackageTwoTwo/SED_IDL_v0.4.tar) Callback: * - Is there a call-back to the Aladin window? I don't see it. Object numbers: - The "Object" numbers are still just SED internal reference numbers, i.e no relation to the catalog IDs. Updating menu selections: (* sort of) - When the SED plot is re-launched from the Aladin Plugins menu, the plot is shown in the default flux[Jy] vs. freq[Hz], but the selected values in the unit menus remain as they were from the previous plot. i.e. the units shown in the menus are not what is in the current plot. Is it possible to update the menus when re-launching? 3. ACE: Tiny window; - At the first launch of ACE in a session, the ACE window is really really tiny, and needs to be stretched. No real problem, but you need to know about it for starting the first time. (Note that default.ace needs to be loaded in the first session) Extraction of catalogs: - Working as normal Photometry modes: - We really only need Ignore, Generic Mag, and Physical Units - The EIS ones are specific for the catalogs in Vizier, but I suggest we just leave it the way it is for the moment. Physical units: - not tested yet 4. Closing the loop: * I've tried to plot the results from ACE in the SED tool. ACE appears to provide all the column names and UCDs required for this to work as set out in the mag_to_flux document. However, it's not working. My notes below include the errors etc., but I gave up when I realised that the test VOTable which worked during testing, also fails. Thomas - can you please track this down? Test 1. ACS F775W image. Generic passbands selected "R" ACE generates a catalog and seesm to have done the right thing. The column names and UCDs seem to be compliant with what was specified in mag_to_flux document... but it doesn't work. The error given is Starting SED... java.lang.NullPointerException at cds.aladin.VOTableEISGenerator.writeCase3Values(VOTableEISGenerator.java:466) at cds.aladin.VOTableEISGenerator.writeData(VOTableEISGenerator.java:285) at cds.aladin.VOTableEISGenerator.writeTable(VOTableEISGenerator.java:653) at cds.aladin.VOTableEISGenerator.writeAllSources(VOTableEISGenerator.java:180) at cds.aladin.VOTableEISGenerator.getOutput(VOTableEISGenerator.java:160) at cds.aladin.Aladin.startSED(Aladin.java:1065) at cds.aladin.Aladin.action(Aladin.java:966) at java.awt.Component.handleEvent(Component.java:5262) at cds.aladin.Aladin.handleEvent(Aladin.java:1829) at java.awt.Component.postEvent(Component.java:3806) at java.awt.Component.postEvent(Component.java:3816) at java.awt.Component.postEvent(Component.java:3816) at java.awt.Component.postEvent(Component.java:3816) at java.awt.Component.postEvent(Component.java:3816) at java.awt.MenuComponent.postEvent(MenuComponent.java:263) at java.awt.MenuComponent.postEvent(MenuComponent.java:263) at java.awt.MenuComponent.dispatchEventImpl(MenuComponent.java:294) at java.awt.MenuComponent.dispatchEvent(MenuComponent.java:273) at java.awt.EventQueue.dispatchEvent(EventQueue.java:447) at java.awt.EventDispatchThread.pumpOneEventForHierarchy(EventDispatchThread.java:191) at java.awt.EventDispatchThread.pumpEventsForHierarchy(EventDispatchThread.java:144) at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:138) at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:130) at java.awt.EventDispatchThread.run(EventDispatchThread.java:98) Starting VOPlot... I didn't know how to get at the VOTable produced by ACE directly, but it is possible to look at it as a vOTable via VOplot. This is included below. It seems to be compliant with our example, so I'm confused why it doesn't work. ?xml version="1.0"?> Object selection from Aladin Ext.Appl object selection from Aladin Using the other modes, EIS EIS x-matched, it still fails. It also fails with our test VOtable ERO3_F850LP_modified.vot.... @
161.0 53.14883