<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7654.12">
<TITLE>RE: [kepler-users] passing multiple data frames between RExpression actors</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>I am in 1.0 on OS X. I don't feel up to installing the dev version from the SVN right now.<BR>
Maybe later when I have more time.<BR>
<BR>
I did figure out a workaround in R that doesn't involve saving the workspace...<BR>
I can pass a list of dataframes.<BR>
<BR>
It's not ideal, since I have to know the name of the list elements I care about<BR>
and check for them manually in the next Rexpression actor, but I think it<BR>
will work for me as a stopgap as it is pretty flexible and seems to work reliably.<BR>
<BR>
Thanks for confirming the bug for me.<BR>
<BR>
Scott<BR>
<BR>
-----Original Message-----<BR>
From: Derik Barseghian [<A HREF="mailto:barseghian@nceas.ucsb.edu">mailto:barseghian@nceas.ucsb.edu</A>]<BR>
Sent: Fri 11/13/2009 2:01 PM<BR>
To: Scott Porter; kepler-users@kepler-project.org<BR>
Subject: Re: [kepler-users] passing multiple data frames between RExpression actors<BR>
<BR>
Hi Scott,<BR>
<BR>
Are you using Kepler 1.0? I tested this with a simple workflow here, <BR>
and it seems to be a bug in 1.0, but not in the current development <BR>
version of Kepler in SVN. (I got two copies of the same dataframe in <BR>
my second Rexpression actor).<BR>
Saving the workspace between Rexpression actors is one solution, but --<BR>
save is a feature we try very hard to avoid using since it subverts <BR>
Kepler -- tokens aren't flowing along relations when it's used, and <BR>
can lead to confusion in later Rexpression actors where it's unclear <BR>
what variables are available. However you might use this as a stopgap <BR>
until we release 2.0, or if you're comfortable with checking out and <BR>
building Kepler from SVN, and willing to deal with things frequently <BR>
breaking and changing as we work towards 2.0, you could try the <BR>
development version of Kepler, the instructions for which are here:<BR>
<A HREF="https://kepler-project.org/developers/teams/build/documentation/build-system-instructions">https://kepler-project.org/developers/teams/build/documentation/build-system-instructions</A><BR>
<BR>
Derik<BR>
<BR>
On Nov 12, 2009, at 11:19 PM, Scott Porter wrote:<BR>
<BR>
><BR>
> I have a workflow where everything works. I can pass a single data <BR>
> frame with no problem between two RExpression actors. However, if I <BR>
> add a port for a second dataframe, the first dataframe port now <BR>
> passes junk data from some random memory location (in some cases a <BR>
> different dataframe, in other cases a random function, etc). Is it <BR>
> a known limitation of the RExpression actor that it can only pass <BR>
> one dataframe? Is the only workaround to pass an R workspace <BR>
> between the actors?<BR>
><BR>
> _______________________________________________<BR>
> Kepler-users mailing list<BR>
> Kepler-users@kepler-project.org<BR>
> <A HREF="http://mercury.nceas.ucsb.edu/kepler/mailman/listinfo/kepler-users">http://mercury.nceas.ucsb.edu/kepler/mailman/listinfo/kepler-users</A><BR>
<BR>
<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>