[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: TARP version 1.1 now available
Michael and all,
Thank you Michael. I am roughly up to 8 million measurement pairs
completed from the new data. Slowly I am getting the bugs out of the three
systems. I have solved the memory card problem with TOM3 by replacing the
memory board. Apparently the memory card would "forget" during the long
time it takes to reset to +50 from the pole. The result was the last image
in the sequence was always garbage. Thus the needed fix below. I put in a
line filter for the chiller, and this reduces the number of bad frames from
chiller turn on. TOM2 controls the chiller. It switches the chiller when
it is not taking data so it does not introduce noise in it's data. But
TOM3 can be making a data transfer when TOM2 switches the chiller on. Thus
it is subject to a noise spike in it's data line. So the fix below is less
needed now, but still a good idea.
"TARP Covers the Sky"
At 10:07 AM 5/16/03 -0400, Stupendous Man wrote:
> Tom Droege suggested that I give a name and version
>to the entire ensemble of software used to reduce
>TASS Mark IV data, so that one might reproduce exactly
>the steps taken in the reduction at some point in the
>future. That's a good idea.
> Therefore, I have created a WWW site for "TARP",
>which can stand for
> TASS Astronomical Reduction Programs
>(or lots of other things). You can find it described
> Today (March 16, 2003), I am announcing version
>1.1 of TARP, which includes
> * XVista 0.1.2
> * bait 0.1
> * match 0.6
> * photom 0.6
> * pipeline 0.4
> Tom has been using TARP 1.0 to reduce data so far.
>The only difference between the two version is in
>the "pipeline" component, which has jumped from 0.3 to 0.4.
>The is no difference in the reductions per se --
>all the numbers (magnitudes, positions, flatfielding, etc.)
>should be exactly the same. Version 0.4 of the pipeline
>scripts are simply a bit more robust to bad data, so that
>they don't halt in the middle of a run. Tom should be able
>to run data through the new version without having to
>intervene manually so frequently.
> Just for fun, I added up the quantity of the software
>in these packages. The total is 62,065 lines of code,
>or about 1.77 MBytes. Phew.
> Michael Richmond