[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20051122: IDV v1.2/v1.3b1 bug: formulas in bundles

>From: Dave Dempsey <address@hidden>
>Organization: San Francisco State
>Keywords: 200511230519.jAN5JG7s021420 IDV derived fields

Hi Dave-

>I've run across a problem generating multi-field displays from saved
>bundles in which one of the fields in the display is generated using a
>formula.
>
>Details: I gave an assignment to my students last week in which I asked
>them to generate the plots described below. All students were running IDV
>v1.2. Some were running on Windows PCs and some were running on Macs.
>
>   (1) Thickness contours and 10 m wind vectors on a color-filled
>sea-level pressure contour plot, using RUC model output; and
>
>   (2) Wind vectors and temperature advection contours on a color-filled
>temperature contour plot at 1000 mb, using the same data source.
>
>
>In the first plot, the 10 mb wind vectors must be generated using a
>formula. In the second plot, the temperature advection must be generated
>using a formula. Both plots initially can be created successfully.
>
>I had the students save each display both as a bundle and as a Quicktime
>movie, both of which are accessible via the Web.
>
>(For each bundle, the data source is accessed via URL.)
>
>Unfortunately, when I use the saved bundles to generate a display (either
>by starting the IDV with the bundle specified as a URL on the command
>line, or by copying the bundle to my favorite bundles directory and
>selecting the bundle from the favorite bundles menu), with one exception
>(noted below) the IDV issues an error message and ends up plotting only
>the two fields in each case that are NOT generated using a formula.

This is a bug that is fixed in the 1.3b1 release.  Thanks for
finding this. The problem occurred because the forumla is used
to define both a system derived quantity (i.e. shows up under the 
Derived tab for gridded data) and as a system formula.  This was
the first time we had actually tried it and unfortunately didn't
test it as you did.

Unfortunately, the only workaround for previous release would be
to create a new formula with a different name, but use the same
formula as the system one.

One other feature we added along with the fix is a labelling 
facility for the the parameters.  With the 1.2 release, the
dialog that pops up for the parameters has columns named
scalar_parameter, D1 and D2.  It's not obvious what D1 and 
D2 should be (but your students figured it out. ;-)).  In
the next release, the D1 and D2 operands will have more
descriptive names (U component, V component).

>I get the same problem when reading the bundles with IDV v1.2 running on a
>PC and v1.2 and v1.3b1 on a Mac (don't know about v1.3b1 on a PC). For the
>second of the two displays above I get the same problem whether the
>bundles were created on a Mac or on a Windows PC.  For the first display
>above, the problem occurs when the bundle is created on a windows PC but
>not when it is created on a Mac.

I thinks the Mac vs. Windows problems are a red herring.  There are
two problems here:

- the bug with the advection field (explains why one of efriebrg's
worked and the other didn't)
- the across bundles are being created with an unknown version
of the IDV.  (explains why neither of these worked) The bundle
says that his was created with version 1.2b2, but I can't
If you can find out exactly what version is being run there, 
we can track that down more.  The problem is that in the 
bundle there is a level set for the 2 m temperatures, but neither
the 1.2b2 release nor the 1.2 release do this.  I seem to remember
that the 1.2RC1 release had a bug in it which might explain
this, but then the version in the bundle should read 1.2RC1.

>To replicate the problem, try running the following:
>
>   Display (1) above:
>
>     (a) Bundle generated on a Windows PC (doesn't work OK on my Mac or
>a Windows PC):
>
>         runIDV \
http://funnel.sfsu.edu/students/across/IDV/DefaultIdv/bundles/MSLP_Thk_SfcWind_
> 05111506_RUC.xidv
>
>(The Quicktime movie version is at:
>http://funnel.sfsu.edu/students/across/www/idv/MSLP_Thk_SfcWind_05111600_RUC.m
> ov)
>
>
>     (b) Bundle generated on a Mac (works OK on my Mac; don't know about a
>Windows PC):
>
>         runIDV \
>http://funnel.sfsu.edu/students/ebriebrg/IDV/DefaultIdv/bundles/MSLP_Thk_SfcWi
> nd_05111600_RUC.xidv
>
>(The Quicktime movie version is at:
>http://funnel.sfsu.edu/students/efriebrg/www/IDV/MSLP_Thk_SfcWind_05111600_RUC
> .mov)
>
>
>   Display (2) above:
>
>      (a) Bundle generated on a Windows PC (doesn't work OK on my Mac or
>a Windows PC):
>
>         runIDV \
>http://funnel.sfsu.edu/students/across/IDV/DefaultIdv/bundles/T_Wind_TAdv_1000
> mb_05111506_RUC.xidv
>
>(The Quicktime movie version is at:
>http://funnel.sfsu.edu/students/across/www/idv/T_Wind_TAdv_1000mb_15111600_RUC
> .mov)
>
>      (b) Bundle generated on a Mac (doesn't work on my Mac; don't know
>about a Windows PC):
>         runIDV \
>http://funnel.sfsu.edu/students/ebriebrg/IDV/DefaultIdv/bundles/T_Wind_TAdv_10
> 00mb_15111600_RUC.xidv
>
>(The Quicktime movie version is at:
>http://funnel.sfsu.edu/students/efriebrg/www/IDV/T_Wind_TAdv_1000mb_15111600_R
> UC.mov)

I'm sorry this exercise didn't work out, because this is exactly
what we had in mind when we created the bundling facility.

Don
NOTE: All email exchanges with Unidata User Support are recorded in the
Unidata inquiry tracking system and then made publicly available
through the web.  If you do not want to have your interactions made
available in this way, you must let us know in each email you send to us.