Difference between revisions of "IVAS Feedback/3.0"

From Northwestern University Center for Atom-Probe Tomography
Jump to navigation Jump to search
(→‎Bugs: submitted feedback)
(→‎Feature Requests: submitted feedback)
Line 76: Line 76:
* A method to insert text in the 3D analysis window
* A method to insert text in the 3D analysis window
* A method to draw basic shapes (the cube/cylinder/spheres (without locked aspect ratios))in the 3D analysis window
* A method to draw basic shapes (the cube/cylinder/spheres (without locked aspect ratios))in the 3D analysis window
=====Root and reconstruction=====
=====Root and reconstruction=====
Line 83: Line 82:
* Make background color of PNGs transparent. --[[User:Karnesky|Rick]] 14:13, 1 February 2007 (CST)
* Make background color of PNGs transparent. --[[User:Karnesky|Rick]] 14:13, 1 February 2007 (CST)

Revision as of 20:40, 2 February 2007

This feedback will be sent to Imago based on demand. Feedback presented at the training meeting has been moved to IVAS Feedback/3.0/training.

Please sign your comments with four tildes (~~~~).

When reporting bugs:

  • Note the machine or machines the the error happened on
  • Steps to reproduce issue
  • Note how often you've been able to reproduce the problem
  • Include a screenshot if it would be useful.
  • If a crash or lock-up:
    • Any other applications running?
    • Data on network or local drive?
    • Save all the files in C:\Documents and Settings\<USERNAME>\Application Data\ivas-3.0\dev\var\log to P:\software\IVAS errors\ under a new directory with a short description of your bug
    • Also put any data files we might use to help reproduce the crash or lookup in this directory.
    • Also put screenshot in this directory.


  • When using the clipping control for a large( z > 1000 nm) dataset, I kept getting "not a valid number" when I tried to manually type in values for the z range. This is even though they were positive integers that were less than the end z value... When using the slider bar I didn't get any error message. Also, if I removed the comma from the fields, I didn't get the error message. This took place on Cain. Mulholla 14:01, 20 November 2006 (CST)
    • Sounds like a legitimate issue. We are receiving more requests to add commas to large numbers. I would assume this wouldn’t trouble you as long as the text parsing handled commas w/o error.
      • Yup.
  • If I have the Display Ranges window open and I create interfaces, run a proxigram, or run the envelope method I am not always able to close the Display Ranges window.
    • We have encountered this issue, the “Display Ranges” window just won’t die. The user can close it but once they come back to the associated 3D display it comes right back
  • On multiple datsets, I cannot change the polygon filter level. I create interfaces, right click on the interface group in the analysis window, and do launch interface analysis. When I then try to change the polygon filter, I get "Unexpected exception" This occurred on multiple datasets on multiple machines(cain and eve). I placed one dataset and the exact error message in P:\software\IVAS errors\3 Mulholla 16:40, 6 December 2006 (CST)
    • Please provide range file/parameters --Rick 11:18, 29 December 2006 (CST)
    • Next time you create interfaces, left click on the interface group node in the Analyses tab. At the bottom of the Analyses tab you'll see the properties of the interface group node. One of the editable properties is the Filter Level (polygons). This filter level is the same as the polygon filter level that you see when you launch the interface analysis panel. The problem you are running into is that the polygon filter level in the interface analysis panel cannot be lower than the Filter Level (polygons) in the properties section of the Analyses tab.
    • When you need to create interfaces and you want to work in the interface analysis tab, first lower the Filter Level (polygons) to the point that you can see all of the interfaces you are interested in. Then launch the interface analysis panel and you will be able to set the polygon filter level to any number as long as it is greater than or equal to the Filter Level (polygons).
      • And thanks to Dave. If that is Mike's problem, I guess we had also talked about it before.

  • When envelope is run in IVAS 3.0 on the “eve” machine, the output text files do not have a .txt identifier. For example, the three output files are:
    • R6_3552 Test Data 08-17-2006 v01.pos.rad
    • R6_3552 Test Data 08-17-2006 v01.pos.log
    • R6_3552 Test Data 08-17-2006 v01.pos.lg.
    • Additionally, when precipitate analysis is run it looks for a file called …pos-lg.txt, which obviously it cannot find. We have to manually add .txt to the end of each file name and also change …pos.lg.txt to pos-leg.txt. This error did not exist in IVAS 2.3.1 Posted by Prakash 10:43, 17 January 2007 (CST)
      • Which computer(s) does this apply to? Have you replaced envelope.exe with the one on ARC? --Rick 10:48, 23 January 2007 (CST)
        • Works fine with my envelope
  • On Cain, clipping stopped working. Can't reliably reproduce. --GH
  • When using average atomic density rather than element specific for a shank angle reconstruction within IVAS 3.0 from root (at Argonne), does not save the reconstruction that is in the preview. It looks like it exports the element specific reconstruction, despite that was not what was selected. Was not able to test if this feature worked in IVAS 2.3.1. Was able to to use root externally from IVAS to get a shank angle reconstruction that uses average atomic density.
    • I was able to export POS file from root outside IVAS but not a root file. Why not?-- Chantal, 13:41, 22 January 2007 (CST)
    • It seems something to do with a limit on the number of mass ranges / element assignment, as I seem to be able to get it work with fewer mass ranges.-- Chantal, 16:51, 22 January 2007
    • OK. So, apparently it is not the number of mass ranges. There is some bug that this feature does not like imported range files, but is OK when you assign the mass ranges by hand in root within IVAS.-- Chantal, 14:46, 24 January 2007
  • When running envelope script on one data, ranges can't be deleted or modified on another data set during reconstruction.--Yang, 2:16pm, Jan 24 2007
  • Image/Movie capture on 3D window and Ctrl-I action sometimes deactivates (especially for graphically complicated data sets). However, the buttons in the main toolbar for these actions continue to work. --Rick 14:13, 1 February 2007 (CST)

Mass >250 amu

So I found this rather interesting. IVAS is supposed to be able to handle mass-to-charge values from 0-300. Apparently when it gets above 250 it starts getting confused and creates a bunch of errors. I got array index errors and general display errors (ie. it couldn't display all and sometimes any of the ranges above 250). So I shifted the data back down below 250 and all is well... Stupid computers. (--Dan)

Feature Requests

Export of Statistical Information

Also, I had a note written down that NU wants to be able to export the statistical information from interfaces. The interface group entry in the analyses panel allows the center, roughness, # of polygons, surface area and volume of all interfaces not currently filtered out to be exported to a csv. Is that sufficient for what the group needs or is there more required? --David Reinhard

Eventually: the total number of enclosed atoms and the composition would be useful.
Also: it should be exportable from wherever some of this information is displayed (so both the interface panel if that is kept & the analysis tree).

API/Scripting and Batch Processing Enhancements

  • An API would be nice
    • It would be nice, but David doesn't know when we'll get one. It doesn't sound like plans to add one have been canceled.
    • Regarding the API. As you might have noticed IVAS has changed quite a bit since 2.3. While the GUI shouldn't see drastic change in the future, the underlying implementation of certain functionality will still be subject to some drastic changes. We are still at a point where we can't publish an API and be certain we are able to maintain it from release to release. Also, we don't have any internal customers who would utilize this feature so it's been very hard for the development team to determine what the proper scope should be. If you could give me an prioritized list of what you would want access to it might help us to start targeting stability in the various APIs we implement and use. Note that I can't commit to delivering anything, but it's never too early to start.

Specific wants for API

Please re-order by your priority!

  • Even minimal documentation (auto-generated javadoc or doxygen would be fine to start with) on all classes which could be accessed from jython would be immensely useful
  • This could be fairly easily expanded so that we'd know what was "stable" and what may be changed
  • For stable and commonly used methods, more helpful/detailed (human-generated) comments would be the nest step
  • Finally, some particularly useful methods may benefit from examples
Precipitate properties
  • A stable way to access the ellipsoid methods used in precip.py
  • A method to get a list of all ROIs and a mehod to return properties on an individual RO (to sort/select them by area/volume/roughness)
  • A method to generate a proxigram on a given ROI
  • A method to export the proxigram to CSV
  • A method to export a POS for a given ROI
3D window interaction
  • A method to insert text in the 3D analysis window
  • A method to draw basic shapes (the cube/cylinder/spheres (without locked aspect ratios))in the 3D analysis window
Root and reconstruction
  • Initial radius calculation, please set minimum voltage to actually be the minimum value in selection box, rather than the box minimum
  • Export of voltage plot, mass history, evaporation rate history, pulse energy history
  • Make background color of PNGs transparent. --Rick 14:13, 1 February 2007 (CST)


Why, when I right-click on 'Mass Spectrum Analysis' and 'Display Bulk Composition' does the total number of ion counts (if I sum up all the individual ion type counts) not equal the 'Spectrum Ranged Ion Count' in the 'Mass Spectrum Analysis' 'Properties' panel?

I am looking at this right now for a straightforward analysis - no complex ions, unknown ion types, or other funny business. The sum of the individual ion counts is ~3.0M, while the 'Spectrum Ranged Ion Count' shows ~2.6M. Why??!!?? --Jason

Are you seeing the bug we reported (which was confirmed):
When the range of an element is changed in the mass spectrum window and the update button on the bottom right is clicked, the Spectrum Ranged Ion Count under properties of Mass Spectrum Analysis is not updated. One has to additionally right click on Mass Spectrum Analysis and then select Update Mass Spectrum. --Rick 10:52, 29 December 2006 (CST)
that's it ... duh. --Jason

Status of 64-bit? --Rick 11:19, 29 December 2006 (CST)

We are very interested in 64 bit platforms. Unfortunately some of our vendor (Java/Sun) supplied libraries are only supplied compiled for 32 bit architectures. This is a problem since if any code in an application (or associated DLLs) is 32 bit compiled the entire application runs in 32 bit compatibility mode (and then subject to the same 32 bit constraints… **sigh**) Fortunately the problem libraries are sources of other headaches (i.e. JMF and movie compression) so there is incentive the replace them independent of the x64 issue. When we do go to x64 we will most likely be skipping XP-64 and going straight to Vista, I have tinkered with XP-64 and found it to be a horrible experience with regards to current application support.
Thanks for the status & we'd appreciate being updated as the libraries are updated or as you migrate away from them.