Difference between revisions of "IVAS Feedback/Pre-2.2"

From Northwestern University Center for Atom-Probe Tomography
Jump to navigation Jump to search
 
 
(One intermediate revision by the same user not shown)
Line 83: Line 83:
 
* <strike>Ability to export movies</strike> ''RK's script or IVAS 2.3''
 
* <strike>Ability to export movies</strike> ''RK's script or IVAS 2.3''
 
* Use a global ivas2 default configuration (for example in c:\documents and settings\all users\ivas2 or in /etc/ivas2) which could store GlobalSettings.py, RootSettings.py, and others).  Still allow users to individually adjust settings (in c:\documents and settings\karnesky\ivas2 or in /home/karnesky).
 
* Use a global ivas2 default configuration (for example in c:\documents and settings\all users\ivas2 or in /etc/ivas2) which could store GlobalSettings.py, RootSettings.py, and others).  Still allow users to individually adjust settings (in c:\documents and settings\karnesky\ivas2 or in /home/karnesky).
[[Category:Tools]]
+
[[Category:IVAS Feedback]]

Latest revision as of 21:37, 3 November 2005

Questions

General

  • Java 1.5 is already on Windows 64-bit & Linux 32/64-bit. Will IVAS run under these? (We use Opterons & would love to be able to use more than 4 GB of RAM in the future).
  • What about ROOT on these alternative platforms?
  • Please provide more info on RHIT and RRAW files, so that we can create simulated data. POS format is already well-understood.

DAVis

Can we "replay" a RHIT for a completed run?
Not from within DAVis. This functionality exists for demo and test purposes, but is not bundled as part of the DAVis release or easily exposed with LAS.
If applying range file later during analysis, can it display the depth profile retroactively (showing history of specimen)?
Specifically no. A more general answer: All binning in DAVis is destructive. Data is received, corresponding histogram bins increments and data tossed.
Can one edit the colors of a range file during an analysis?
No, color handling is archaic. Colors are assigned to ranges in the order in which they are read. The color list is fixed, assignment within this list could be modified by rearranging the order of ranges in the source range file.

ISDB

How do you use exported XML files? Can we bring them into excel/openoffice? Why not CSV? (Work-around
use XSLT sheet Rick made to convert the XML file to CSV.):XML files can be opened with Excel. They are easier to open than CSV (not prompted with Text Import Wizard) and there's better record sorting built into the column headers.
  • Which version of Excel on which platform? We can't open this in Excel some version(s) on some platform(s) or in gnumeric 1.4.1 (Linux) or on OO.o calc 1.1.4 (Linux). Furthermore, shouldn't sorting be dependent on the spreadsheet * NOT the table?
Is there a way for us to record which user was the PI for a particular run (user is always "admin")
Add Operators with the ISdb application via the Configuration dialog. These operators will appear in a drop down list as users in the Experiment Summary dialog at the end of an acquisition.

ROOT

What are the Analyzed Data, Live Data, Live Calibration, and RHIT Archive directories each for?
The "Analyzed Data" directory is the location that analysis files ("root files" get written to on File->Save Analyzed)
If "Live Data" directory is a valid location, the File->Open Current RHIT menu item appears - this opens the most recent .RHIT file in this directory. This is intended for use on the acquisition client to easily open the current (or most recent) file
If "Live Calibration" directory is a valid location, the File->Save Calib... menu item appears: This opens a file dialog initialized to the Live calibration directory.
Edit->Preferences. What is Max Bowl Subdivide?
The bowl correction currently is based on subdividing the detector area into NxN regions, each with a quadratic surface fit. For N=1, a single polynomial surface is used - 6 parameters. N=2 gives 4 polynomials, etc. As N increases, the number of fit parameters increases (number = 6*N**2) and the number of degrees of freedom decrease. At the end of each bowl iteration the fit results are showns as:
SUBDIVIDE = 1 chi2/dof = 2.197
SUBDIVIDE = 2 chi2/dof = 2.009
SUBDIVIDE = 3 chi2/dof = 1.887
SUBDIVIDE = 4 chi2/dof = 1.846
CBowl::Compute() best correction, number of params = 96
The subivision choice with the smallest chi2/degree of freedom is chosen. Sometimes, low statistics runs or runs that do not populate a large enough area of the detector or when small fiducial volumes are selected, the N=4 is too large and the fits fail or become unphysical - check the bowl correction isochrones - Diagnostics->Bowl() to make sure the correction is really bowl shaped.
Edit->Preferences. What is DeltaTOF Cut?
This is a quality cut: the maximum difference, in sigmas, between that the x- and y- TOF measurement can have: IF these measurements differ by more than this amount, the ion measurement is eliminated. The 1 sigma value (in nsec) can be read off of the View->Delta TOF plot: 1 sigma = FWHM value divided by 2.35 for a gaussian - approximatly the RMS value displayed in the stats box. Note: This histogram has 1 bin per TDC lsb.
What do n peaks do in mass calibration? 1 for shifts, 2 for shifts/scaling, 3 for more stats?
Presently the mass calibration assumes that mass = k*TOF^2. When n=1 is used, only k is determined. When n>1 is choosen, k and a T0 shift (e.g. TOF-->TOF-delta) are both fit for as long as the the leap.ini parameter [MassCalib]T0 is set to "Fit". If this parameter is numeric, that value is used for T0 and T0 shifts are not fit for.
How are peaks weighted during mass calibration? Are stronger peaks more heavily weighted?
The error on the peak's mean used to be used - there is a bug and no weights are used. This will be fixed in 2.1 release.
Is it important to select peaks at both sides of the spectrum or peaks that are both very high and very low?
In general, it is best to select as many peaks as possible. Also, the broader the AMU span, the better the fit. In addition to fixing problems (like the bug above), we are currently working on improving the mass calibration model.
Viewing the delta TOF plot on log scale (zoomed out); why is it not symmetric?
There are systematic differences in the delta TOF accross the detector - the origin is not entirely understood. The Diagnostics->TOF Map display shows how delta TOF varies across the detector. The units on the z-axis are in TDC units - about 30 psec each
Can one export 'header' as CSV?
Not currently. We can possibly provide something in the future.
Are View->Delta TOF plot and Diagnostics->delta TOF vs. XY plots correlated?
The View->Delta TOF plot is histogram of all delta TOF values. the Diagnostic->delta TOF vs. XY plot is the average value of the delta TOF as a function of detector position in mm. There is also a low-res version of the latter plot, Diagnostic->Delta TOF Correction (detector position in raw TDC units, detector-averaged TOF value is subtracted) This histogram is used to generate the Diagnostics->Corrected Delta TOF plot - this histogram is View->Delta TOF with the spatial variation removed. In a future release of root, this will be value used for the delta-TOF quality cut.

IVAS

What does "Get System Script" do?
It copies the selected script(s) into the user script directory. Because of the search path (see the Scripts page in the on-line help) you should rename the file, especially if you're going to modify it.
What's the difference between POS file count and Histogram count (under 'Details'); are there dropped ions?
Yes, the historgram goes from 0-250 AMU and there are underflows and ions with m/q's greater then 250 in the POS file.

Bugs

General

  • At. vol. seems to be rounded strangely: 5 sig figs in IVAS. 4 in root. Sometimes rounded correctly, sometimes rounded down.

DAVis

  • Evaporation rate is more accurately described as a detection rate since it is associated with a field of view

ROOT

  • ROOT crashes in the middle of mass calibration with 11 selected peaks, 50 million atoms data set.
  • 'efficiencys' is spelled wrong
  • In the Range File Editor, 'Refresh' doesn't update the display

IVAS

  • "Aspect" is extremely misleading and confusing. It would be more intuitive if one could rather change the aspect ratio of a set (so that all datasets were imported with a displayed aspect ratio of 1:1:1 & one could easily return to this ratio without having to remember the initial "aspect" of the box)
  • Add System Script to Toolbar requires a restart Imago unable to reproduce this
  • GlobalSettings.py and RootSettings.py don't seem to be generated for each new user. These are useful because we want EnableOpenPos to be 1, which doesn't seem to be the default (it should be!) & we need to change the directories used by root.
Changing a range (by Rt. clicking in the range area to the left of 3-D reconstruction) resets the at. vol. to 0
Range colors (as seen in range file manager and analysis window) are not remembered after restarting IVAS; default set of colors takes over.
Add a 'user-startup' script (which is run when IVAS starts up) in your user scripts directory that include as many ColorSpec.setColor() calls as you'd like: e.g.
ColorSpec.setColor("Ni",0,255,0)
These can be taken from a saved analysis script that has the desired range colors.

Requests

General

  • Associate Imago's file extensions with correct programs under windows (so we can immediately open IVAS or ROOT by selecting a file)

ISDB

  • Export tables to CSV in addition to XML

ROOT

  • A manual on how to use the command line interface of ROOT
  • logging of command window
  • export 'header' as CSV
  • Autosave an analysis. Either every so many minutes or at least after time-intensive steps
  • Get Ion List should default to :1, so that we can more quickly add single elements (which are so common)
  • ZX/ZY plots in the analysis should have consistent axes (so we can visually judge how good a reconstruction is by aspect ratio)

IVAS

  • Key combination while click & dragging to rotate 3-D reconstruction in plane of screen
  • Proxigram
  • Documentation on ALL system scripts
  • Ability to set an alternate text editor, as with the PDF-viewer
  • Ability to change the ticks used in the m/q histogram
  • Ability to tab in the mass range window
  • Filesystem Browser should automatically select the directory based on current directory if a directory isn't selected
  • Ability to apply statistics to isoconcentration surfaces (e.g., transfer functions) in order to get smoother, more-contiguous surfaces
  • Ability to export movies RK's script or IVAS 2.3
  • Use a global ivas2 default configuration (for example in c:\documents and settings\all users\ivas2 or in /etc/ivas2) which could store GlobalSettings.py, RootSettings.py, and others). Still allow users to individually adjust settings (in c:\documents and settings\karnesky\ivas2 or in /home/karnesky).