|
Spatial Analysis and Decision Assistance
Bugs and other "Features" in SADA 4.1.50
Ok, a bit humiliating, but here is a list of errs (just kidding its spelled errors) encountered with 4.1.50. We will try and release a version soon that corrects some of the more
obnoxious problems. In the meantime, we want users to be aware of what is going on. Please be patient as we strive to make
SADA a better product. If you have any other known problems, please send us an
email
being as descriptive as possible concerning the nature of the error and the events that led up to the error.
We are planning a release of SADA on June 7th, 2000 (4.1.60) that will correct the mass volume and the MARSSIM errors listed below.
SADA doesn't work with European decimal = , setting
You will have to set your regional settings such that decimal = "." Please see tech support.
This manifests itself in all sorts of weird ways. If you're on the other side of the pond, make sure you set this first. This likely
will not get fixed unless we get some Euros from our good friends over there. (hint, hint)
SADA Mass and Volume calculation are wrong when density <> 1
This is a serious mistake and we want users to be aware of it. If you use a density parameter greater than 1 in your set decision criteria step,
SADA will report erroneous volume and mass numbers. The spatial extent of the area of concern is still correct, the number of blocks to remediate will still be
correct but both the volume and mass will be wrong. If you use a density = 1 (default), no worries. If density <> 1 then what really
happens is your volume estimate is really your mass estimate and your mass estimate is just wrong. In particular, suppose you only had to clean
1 block that was 1ft x 1ft x 1ft. Then
density =1, number of blocks to clean = 1, volume = 1, mass = 1 (this is correct)
density =2, number of blocks to clean = 2, volume = 2, mass = 4 (this is wrong)
For now, always use density of 1. If you need to estimate mass, then in SADA the mass is calculated (when done correctly) by just multiplying
your density parameter times your volume. Do this outside SADA.
MARSSIM Module Step 4 WRS Test
In the SADA MARSSIM assessment, the Wilcoxon Rank Sum (WRS) test implemented by SADA under some circumstances contains an error.
Specifically, SADA compares survey data to background rather than comparing survey data minus background to the DCGL when the WRS test is implemented. This is a later step in the MARSSIM process and could result in a survey unit being flagged as containing residual radioactivity in excess of the release criterion when a proper implementation of MARSSIM would show that the site does not contain residual radioactivity. This error will apply only to users with a survey site that meet the following conditions:
The survey is for radionuclides in the survey unit that are associated with a background data set (hence the need for a WRS test instead of the Sign Test),
The site contains survey concentrations that are above the DCGLw,
The maximum of the survey concentrations minus the minimum of the background exceeds the DCGLw,
The mean of the survey concentrations minus the mean of the background is less than the DCGLw,
A WRS test indicates that the survey data exceeds background *and* a WRS test of the survey data minus background versus the DCGLw does not show an exceedance,
Elevated Measurement comparisons do not show an exceedance of the DCGLemc at the survey unit.
Under these circumstances, SADA may classify a site as containing residual radioactivity when the MARSSIM process should say otherwise.
While we believe that the possibility of a site producing data that would fit the above criteria is very small, we recommend that SADA/MARSSIM users review any site assessment where these circumstances apply.
SADA crashes while moving judgmental sampling points
This is our famous "ghost" error. We haven't been able reproduce it in a development environment. But we know its real because it always happens
to us in TRAINING! In V5, the entire internal managment of judgmental design points is changing so perhaps this will "exorcise" our ghost bug.
Switching to Pooled Data and then back again and i lose most of my interview options
Sometimes, when you switch to Pooled data, SADA correctly limits the number of options in the interview box to those things
you can do over all contaminants given the analysis you currently have selected. But then when you switch back to a single
contaminant, SADA doesn't bring all those nice options back. Truth is SADA gets stuck and you have to rattle its cage a bit.
to do this, just toggle the media or the analysis and it comes back. Fortunately, no errors are created just bewilderment.
Help file and Vista....
If you opened SADA in Microsoft Vista and tried to use the help file you know it doesn't work. Well both you and SADA are not alone.
Apparently Microsoft has elected to no longer directly support all former help file formats. We've checked into it and
some of the big help file software creaters like Robohelp are scrambling to respond to this. In the meantime, Vista offers
a way to get some fix online and make it work. At this point, that's all we know to do.
ROSE diagram will not export to a csv
With a rose or variogram map up, the export to csv file doesn't export anything.
Data editor buttons don't move when you resize the page
On the data import data editor page. Certain buttons don't move their position when you change the size of the form.
Labels on data plots don't format according to Legend Format
The formating of "plotted values" doesn't respond to or isn't in sync with the legend format.
Sometimes when I go to legend manager, its empty and asks me to add a new one. Then it dies.
Yep. We know. This is another famous ghost error, but we think we caugt it in time for V5.
This can't be all there is!!
Probably not - but that's what we know about now. Sometimes things happen that are unexpected in an analysis but technically aren't bugs.
Particularly, in the geospatial analysis where numerical methods and approaches might be used inappropriately. For these kinds of things
check out our tech support.
|
