Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362012-06-01T17:52:52ZEcoinformatics Redmine
Redmine Bug #5612 (New): String index out of range in R actorhttps://projects.ecoinformatics.org/ecoinfo/issues/56122012-06-01T17:52:52ZClaas-Thido Pfaffsabsirro@arcor.de
<p>The R actor can only hand down a certain size of string array through a port. <br />I have attached a workflow which shows that. To trigger the exception only modify the first actor. The first actor also contains more description.</p>
<p>Best regards Claas</p> Bug #5593 (New): use setDefaultWidth in actorshttps://projects.ecoinformatics.org/ecoinfo/issues/55932012-04-28T22:05:50ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Building workflows in Kepler these days I spend a lot of time changing relation widths from Auto to 1, because w/ Auto the following error message occurs:</p>
<p>--<br />ptolemy.kernel.util.IllegalActionException: Error recording workflow structure: Error getting port width: The width of relation .DataTurbineTimeSlice.relation can not be uniquely inferred.<br />--</p>
<p>The error includes a suggestion, it sounds like we should take a pass through all actors and update code to call setDefaultWidth as appropriate:</p>
<p>--<br />Please make the width inference deterministic by explicitly specifying the width of this relation. In the user interface, right click on the relation, select Configure and change the width. Note that some actors may need to have their Java code updated to call setDefaultWidth(1) on the output port.<br />---</p> Bug #5590 (New): RExpression forces pdf if 'automatically show graphics' is checkedhttps://projects.ecoinformatics.org/ecoinfo/issues/55902012-04-26T01:34:05ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Look into this forced format change side-effect which is confusing to a user. If you select e.g. png, and 'auto show..', you get a pdf.</p> Bug #5269 (New): Selectordialog and Case actors do not work well togetherhttps://projects.ecoinformatics.org/ecoinfo/issues/52692011-01-07T00:54:25ZRobert Weberrweber221@comcast.net
<p>The Selectordialog actor in Kepler outputs a strange sort of string array that cannot be cast or converted into a string scalar. The control port of the Case actor expects a string scalar to compare against the names of the "refinements". Therefore, these two actors, which look like they should work in tandem, fail to collaborate. The attached example shows the problem. I've tried a number of fixes none works;</p>
<p>typing the output port of Selectordialog<br />typing the control port of Case<br />using the Expression actor to recast the output of Selectordialog</p> Bug #4881 (New): RExpression outputs [false] for any matrix of booleanshttps://projects.ecoinformatics.org/ecoinfo/issues/48812010-03-12T18:53:40ZOliver Soongsoong@nceas.ucsb.edu
<p>In an RExpression, add an output port "out" and use this R script:</p>
<p>out <- matrix(c(TRUE, FALSE, TRUE, FALSE), 2)</p>
<p>I get [false] as the output token, which seems to be because RExpression.java on line 1108 says:</p>
<p>mt = new BooleanMatrixToken();</p>
<p>instead of:</p>
<p>mt = new BooleanMatrixToken(valS);</p>
<p>or something similar. I'm not actually sure such a constructor exists, but it seems like it should given the other constructors in the lines before.</p> Bug #4880 (New): RExpression cannot output matrix of stringshttps://projects.ecoinformatics.org/ecoinfo/issues/48802010-03-12T18:45:03ZOliver Soongsoong@nceas.ucsb.edu
<p>Create an RExpression with the output port "out" and for the script, use:</p>
<p>out <- matrix(letters, 2)</p>
<p>Looking at the code, it seems as if a matrix of strings was never really considered. Commas in the strings would likely cause problems, and I don't actually know if there's a StringMatrixToken.</p> Bug #4879 (New): RExpression replaces ) in string vectors with } when outputting to a porthttps://projects.ecoinformatics.org/ecoinfo/issues/48792010-03-12T18:37:49ZOliver Soongsoong@nceas.ucsb.edu
<p>In an RExpression, create the output port "out" and in the R script, use:</p>
<p>out <- c("(", ")")</p>
<p>Hooking this to a Display actor, I get {"(", "}"}. This is because RExpression.java line 1139 (I think) has:</p>
<p>temp = temp.replace(')', '}');</p>
<p>which replaces all occurrences of ), even within the string.</p> Bug #4876 (New): RExpression - recursively handle arrayshttps://projects.ecoinformatics.org/ecoinfo/issues/48762010-03-11T23:25:19Zben leinfelderleinfelder@nceas.ucsb.edu
<p>Currently you can only have a list of lists - it would be nice to handle the Token->R->Token conversion recursively so that there was no limit to 'n'</p> Bug #4827 (New): RExpression does not handle matrix tokens correctly.https://projects.ecoinformatics.org/ecoinfo/issues/48272010-02-23T20:38:25ZTom Parrisparris@isciences.com
<p>I've attached a simple workflow that demonstrates a bug in the way the RExpression actor translates Kepler/Ptolemy matrices to R matrices. There seems to be a problem in the autmatically generated "matrix" statement in R that is likely due to confusing row and column counts.</p>
<p>A simple 2 row x 5 column constant matrix ([1, 2, 3, 4, 5; 101, 102, 103, 104, 105]) is passed to Rexpression on the min port. The R script simply displays the matrix and passes it along to the mout port.</p>
<p>There are three displays.</p>
<p>1. min displays the matrix before it enters the RExpression actor. It is fine.</p>
<p>2. mout displays the matrix after it leaves the RExpression actor. It too is fine (indicating that the translation problem inverts itself when baorcasting to the output port).</p>
<p>3. The problem is in the Routput display.</p>
<blockquote>
<p>min</p>
</blockquote>
[,1] [,2] [,3] [,4] [,5]<br />[1,] 1 3 5 102 104<br />[2,] 2 4 101 103 105
<p>Note how the matrix is scrambled. The automatically generated R script line that reads:</p>
<pre><code>`min` <- matrix(c(1, 2, 3, 4, 5, 101, 102, 103, 104, 105), nrow=2,ncol=5)</code></pre>
<p>should read</p>
<pre><code>`min` <- matrix(c(1, 101, 2, 102, 3, 103, 4, 104, 5, 105), nrow=2,ncol=5)</code></pre> Bug #4407 (New): RExpression problems with named non-numeric vectorshttps://projects.ecoinformatics.org/ecoinfo/issues/44072009-09-24T15:12:50ZOliver Soongsoong@nceas.ucsb.edu
<p>RExpression seems to have a problem outputting named non-numeric vectors that is not present with named numeric vectors.</p>
<p>See attached workflow for examples.</p> Bug #4358 (New): variable setter seems to lock Kepler in the wrapping up stage of a workflowhttps://projects.ecoinformatics.org/ecoinfo/issues/43582009-08-31T22:15:15ZOliver Soongsoong@nceas.ucsb.edu
<p>Running the attached workflow seems to cause Kepler to freeze at Wrapping Up. I'm not sure if I just need to wait (for a really long time) or whether there's some other issue.</p> Bug #4268 (New): RExpression output ports search entire search path for objects to exporthttps://projects.ecoinformatics.org/ecoinfo/issues/42682009-07-27T19:46:09ZOliver Soongsoong@nceas.ucsb.edu
<p>This is debatable, but currently the RExpression actor will search the entire search path for an object to export to an output port. For example, if I create an output port "data", but do not assign it a value, the RExpression actor will export the function data from the utils package. An alternative would be to limit exports to variables in the global environment (.GlobalEnv). That would basically mean changing the export commands (file, serialize, close) in an if(exists("data", .GlobalEnv, inherits = FALSE)) statement.</p> Bug #4168 (New): User documentation creates bloat in workflowshttps://projects.ecoinformatics.org/ecoinfo/issues/41682009-06-18T16:49:11ZChad Berkleyberkley@nceas.ucsb.edu
<p>Currently, the user documentation is copied into every actor instance as the actor is drug to the canvas. The problem that this is addressing is that people need to be able to customize the documentation for actors that they have customized on the canvas. The problem with this approach is that it creates unnecessary duplicate documentation in every actor instance. One solution would be to diff the documentation against the original and just store what is different in the actor instance.</p> Bug #4057 (New): UserInteractiveShell messagehttps://projects.ecoinformatics.org/ecoinfo/issues/40572009-05-08T14:46:03ZChris Weedchrisweed@gmail.com
<p>The message in the UserInteractiveShell doesn't make sense, and gives a "null" at the beginning.</p>
<p>Here is the current message:<br />The outputs of the previous step.<br />Please double click on your selections to be sent to the next step</p>
<p>This seems to come from DoubleShellTextAreaPanel.</p>
<p>I would expect it just to print my input message.</p> Bug #4041 (New): synchronization in matlab actorhttps://projects.ecoinformatics.org/ecoinfo/issues/40412009-04-30T17:28:06ZDaniel Crawldanielcrawl@gmail.com
<p>The matlab actor starts a new matlab process in preinitialize and closes it in wrapup. Each instance of the actor in a model starts and stops a separate matlab process. However, they all synchronize on a shared object, which prevents concurrent matlab execution.</p>
<p>Ideally, the actors should be able to run concurrently. Why are they synchronized? (Maybe the matlab engine shared library is not thread-safe)?</p>
<p>If they cannot run concurrently, then only one matlab process should be used per JVM.</p>