[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Swarm-Support] New code: Interpolator, time series input manager
From: |
Steve Railsback |
Subject: |
Re: [Swarm-Support] New code: Interpolator, time series input manager |
Date: |
Sat, 17 May 2003 13:10:28 -0700 |
"Paul E. Johnson" wrote:
>
> Dear Steve:
>
> Thanks very much for posting it. I've been studying the download of the
> time series manager and have a couple of questions.
>
> 1. THis may be an obvious one, but:
>
> This class assumes that only one variable is loaded from each file?
Yes...My basic philosophy was to keep the Swarm code simple and do the
data preparation in a spreadsheet. So- nobody cares about file size any
more; it's easy in a spreadsheet to crank out a separate file for each
variable, and it makes the TimeSeriesInputManager easy to use in your
Swarm model- just create one instance of it for each input variable.
> 2. In the download, there is a mysterious file called
> TimeSeriesInputManagerP.h
It's a protocol. If you want to know why we included a protocol, there
is a simple good answer, and I'm sure Steve Jackson knows it.
> 3. Swarm can write and read hdf5 files. Did you ever experiment with
> that as a way to manage time series data? I have succeeded with the
> output of hdf5, but not the reading-in part. But I did not try with any
> vigor at all.
No- again, our idea was that you almost always end up messing around
with the data in a spreadsheet to clean it up, check it, look for
missing or bad values, etc.; so then make it as easy as possible to get
it into your model from the spreadsheet.
Steve
--
Lang Railsback & Assoc.
250 California Ave.
Arcata CA USA 95521
707-822-0453; fax 822-1868