Menu
▾
▴
oscill8-devel — list for the developers
You can subscribe to this list here.
| 2005 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
(3) |
Jul
|
Aug
|
Sep
(4) |
Oct
(3) |
Nov
(1) |
Dec
|
|---|
|
From: Emery C. <ec...@vt...> - 2005-11-18 16:22:50
|
Frank, I notice that SimDriver plots and labels data according to the order in which it found the species IN THE SBML file, not in the order that the Simulator says they are in... that's not GOOD! A simulator may, for reasons of speed or whatever, decide to order things it's own way. In the case of Oscill8, the XPPTranslator outputs things in a different order than that found in the SBML file AND THIS IS NECESSARY for speed, since floating species and dependent species are handled as two separate, CONTIGUOUS blocks of memory, and therefore increases speed when copying data etc. Can this be fixed? Also, you mentioned that the Oscill8 simulator sometimes returns more data than was requested... 1001 instead of 1000. This is correct, and is especially necessary when we model with events. The simulator needs the flexibility to return more data to reflect the shape of the graph (at discrete events there is usually a huge jump, and it's very helpful to have extra data points to do this). The other alternative is to bail out when you've got enough points, BUT THAT is much worse from a user standpoint, since that would stop the simulation BEFORE reaching the end time. One last question: What methods must I implement to have the continuous timecourse in the SimDriver work? BTW, just about done with testing new Oscill8; I'll upload it shortly. Emery |
|
From: Emery C. <em...@os...> - 2005-10-17 22:53:53
|
All, I didn't do the merge properly... I'll fix this tomorrow. In any case, there is the latest installer up at sourceforge, v1.13. Emery |
|
From: Emery C. <ec...@vt...> - 2005-10-17 18:51:16
|
FYI, I'm moving the oscill8_bsam branch back into the HEAD rev. today... after this, if you care to do any development, you'll need to move your working copy back to the main branch to keep in sync. Will done by 5pm EST. Emery |
|
From: Emery C. <ec...@vt...> - 2005-10-03 06:17:49
|
Nick, I spent about three hours tonight getting rid of 95% of the errors you detect with the test suite... of the remaining fails, only ONE of them is an actual problem--though it's questionable who's to blame! (The arcsin test bombs because of a rounding error on 1.0--if you run the test with the initial value at 0.9999999, it works fine...). The rest of the remaining fails are simply unimplemented features, like "if/then/else" or "&" or "|" or blah blah... no crashes or hangs anymore on anything; I've at least disallowed the syntax for these tests (if you use the -l option of o8core, you'd get the error messages in a log file). You can get the new version at the usual place, v1.10.1. Enjoy, Emery |
|
From: Emery C. <ec...@vt...> - 2005-09-28 20:31:28
|
Herbert, By events, do you mean discrete events during the time course? If so, then yes, it does. (If you mean windows events, then no... it's just a simple command line service utility). To do events, you use the global flag as in XPP, but with restrictions. See the Oscill8 documentation for the syntax. See ya, Emery Herbert Sauro wrote: > I don't suppose o8core knows how to handle events? > > Herbert > > -----Original Message----- > From: Emery Conrad [mailto:ec...@vt...] > Sent: Tuesday, September 27, 2005 2:06 PM > To: Nick Allen > Cc: osc...@li... > Subject: [Oscill8-devel] oscill8 is ready for you > > Nick, > > Oscill8 should be ready for you now. Here's a sample command-line > invocation of the core that will mimic xppaut (note that you are > explicitly giving o8core command line options for the time series run): > > > o8core.exe -m yeast.ode -ts:out out.txt -ts:t_end 2500 -ts:n_points > 500 -ts:format xpp -v 1 > > NOTE: it's pretty important that you set -v 1 (otherwise you'll get VERY > verbose output). > > ALSO: We should talk about how to streamline things if you're going to > do multiple calls to o8core for many, many time series... the > initialization and run and data write take about 2.2 seconds on my > laptop for the budding yeast model), but the actual solver run itself > takes less than about 3/4 second or so..., so you stand to gain a lot if > we do a simple socket connection or something (which is what I already > have implemented for others). > > Emery > > > > > > > > ------------------------------------------------------- > This SF.Net email is sponsored by: > Power Architecture Resource Center: Free content, downloads, > discussions, > and more. http://solutions.newsforge.com/ibmarch.tmpl > _______________________________________________ > Oscill8-devel mailing list > Osc...@li... > https://lists.sourceforge.net/lists/listinfo/oscill8-devel > > > ------------------------------------------------------- > This SF.Net email is sponsored by: > Power Architecture Resource Center: Free content, downloads, discussions, > and more. http://solutions.newsforge.com/ibmarch.tmpl > _______________________________________________ > Oscill8-devel mailing list > Osc...@li... > https://lists.sourceforge.net/lists/listinfo/oscill8-devel > > |
|
From: Herbert S. <Her...@kg...> - 2005-09-27 21:37:26
|
I don't suppose o8core knows how to handle events? Herbert=20 -----Original Message----- From: Emery Conrad [mailto:ec...@vt...]=20 Sent: Tuesday, September 27, 2005 2:06 PM To: Nick Allen Cc: osc...@li... Subject: [Oscill8-devel] oscill8 is ready for you Nick, Oscill8 should be ready for you now. Here's a sample command-line invocation of the core that will mimic xppaut (note that you are explicitly giving o8core command line options for the time series run): > o8core.exe -m yeast.ode -ts:out out.txt -ts:t_end 2500 -ts:n_points 500 -ts:format xpp -v 1 NOTE: it's pretty important that you set -v 1 (otherwise you'll get VERY verbose output). ALSO: We should talk about how to streamline things if you're going to do multiple calls to o8core for many, many time series... the initialization and run and data write take about 2.2 seconds on my laptop for the budding yeast model), but the actual solver run itself takes less than about 3/4 second or so..., so you stand to gain a lot if we do a simple socket connection or something (which is what I already have implemented for others). Emery ------------------------------------------------------- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl _______________________________________________ Oscill8-devel mailing list Osc...@li... https://lists.sourceforge.net/lists/listinfo/oscill8-devel |
|
From: Emery C. <ec...@vt...> - 2005-09-27 21:06:20
|
Nick, Oscill8 should be ready for you now. Here's a sample command-line invocation of the core that will mimic xppaut (note that you are explicitly giving o8core command line options for the time series run): > o8core.exe -m yeast.ode -ts:out out.txt -ts:t_end 2500 -ts:n_points 500 -ts:format xpp -v 1 NOTE: it's pretty important that you set -v 1 (otherwise you'll get VERY verbose output). ALSO: We should talk about how to streamline things if you're going to do multiple calls to o8core for many, many time series... the initialization and run and data write take about 2.2 seconds on my laptop for the budding yeast model), but the actual solver run itself takes less than about 3/4 second or so..., so you stand to gain a lot if we do a simple socket connection or something (which is what I already have implemented for others). Emery |
|
From: Emery C. <ec...@vt...> - 2005-09-01 14:10:17
|
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Vijay,<br>
<br>
Update to the latest Oscill8, v1.9.23. It has better default values set
for the Sample Period/Min/Max. Tersely, here are<br>
what the parameters mean:<br>
<br>
double period_min; //!< min period to look for<br>
double period_max; //!< max period to look for<br>
double h_min; //!< min step for ODE solver (and use
this close to finding period for accuracy)<br>
double h_max; //!< max step for ODE solver (when we
think we can make big jumps)<br>
double h_init; //!< initial step for ODE solver<br>
double x_match_tol; //!< (xi-xi_at_last_period) <
match_tol ==> match<br>
double period_match_tol; //!< (peri-peri_at_last_match) <
period_tol ==> period is detected match<br>
int period_match_max; //!< (peri-peri_at_last_match) <
period_tol ==> period is detected match<br>
int p_index; //!< index of parameter to continue<br>
double dp; //!< sample step for parameter in
question<br>
int n_max; //!< maximum number of valid (positive)
period calculations before quit<br>
int edge_max; //!< number of bad (negative) results
before quit<br>
int x_index; //!< variable on which to calculate
period<br>
<br>
In answer to your questions below:<br>
<br>
1) when you say "steady states", this refers to a 1 parameter
continuation. You would set the parameter range to .2 to 1.2 and run a
continuation (parameter ranges are set through the "Set Model Data"
button in the Run Config dialog).<br>
<br>
BUT, do you mean the max and min of the oscillation? If so, then you
want to start with a parameter value in the .2 to 1.2 range, and choose
n_max large and dp appropriately (to get a smooth number of samples).
The sampling of the period/min/max starts from your initial value and
goes as FAR as it can in each direction before it can't calculate the
period "edge_max" times in a row. This should take you to the
boundaries of your oscillatory domain.<br>
<br>
2) When the graph shows nothing, it wasn't able to calculate the period
at ALL.. this mean either 1) you didn't set the parameter for the
sampling well or 2) you forgot to put the initial parameter value into
the oscillatory domain.<br>
<br>
3) More information on a HB is in the works; but, this is not something
that AUTO provides, so I haven't yet added this feature. It's on my
list.<br>
<br>
Don't every worry about asking questions!!<br>
<br>
Emery<br>
<br>
<br>
<br>
Vijay Chickarmane wrote:
<blockquote
cite="mid...@de..."
type="cite">
<meta http-equiv="Content-Type" content="text/html; ">
<meta content="MSHTML 6.00.2800.1515" name="GENERATOR">
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">Hi Emory,</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">Sorry for the delay, I did set up the x_index as
2 and I was</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">able to obtain the limits. But I have no feel
for what the </font></span></div>
<div><span class="182293416-31082005"></span><span
class="182293416-31082005"><font color="#0000ff" face="Arial" size="2">parameters
mean. Could you tell me how I can do the foll.</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">1. If I want to get the steady states for a
system with the</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2"> parameter ranging from .2 to 1.2, how could I
set this up.</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">2. Sometimes the program cannot compute the
steady states,</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">the graph does not show anything, usually I
reset the parameter</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">and try again. But since there is no error
message, one does not</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">know what is hapenning.</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">3. Can we get more information on a HB, ex: sub
or supercritical?</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">Hope this deluge of questions will not take up
too much of your time.</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">thanks for the help,</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">best,</font></span></div>
<div><span class="182293416-31082005"><font color="#0000ff"
face="Arial" size="2">vijay</font></span></div>
<div><span class="182293416-31082005"></span> </div>
<div><span class="182293416-31082005"></span> </div>
<blockquote style="margin-right: 0px;" dir="ltr">
<div class="OutlookMessageHeader" align="left" dir="ltr"><font
face="Tahoma" size="2">-----Original Message-----<br>
<b>From:</b> Emery Conrad [<a class="moz-txt-link-freetext" href="mailto:ec...@vt...">mailto:ec...@vt...</a>]<br>
<b>Sent:</b> Friday, August 26, 2005 11:19 AM<br>
<b>To:</b> Vijay Chickarmane<br>
<b>Subject:</b> Re: Issues with Oscill8<br>
<br>
</font></div>
Vijay,<br>
<br>
Which parameters where you interested in the most? the sampling works
fine for me for some of the parameters, but you need to choose your
algorithmic pars correctly (for example, x_index should be set to 1 or
2 given that x0 is a very "stiff" variable).<br>
<br>
Let me know which parameters you're interested in and I'll tell you how
to set the algorithm to run correctly.<br>
<br>
Emery<br>
<br>
<br>
<br>
Vijay Chickarmane wrote:
<blockquote
cite="mid...@de..."
type="cite">
<pre wrap="">Hi Emory,
Greetings, I am sending you a model
which has oscillatory behavior, but I was
unable to obtain the max/min of the oscillations.
It could very well be that I did not use the
tool properly. Could you let me know how
you are able to use this feature.
best wishes,
vijay
-----Original Message-----
From: Emery Conrad [<a class="moz-txt-link-freetext"
href="mailto:ec...@vt...">mailto:ec...@vt...</a>]
Sent: Wednesday, August 24, 2005 12:34 PM
To: Herbert Sauro
Cc: Vijay Chickarmane
Subject: Re: Issues with Oscill8
Vijay and Herb,
Sure thing... the period/min/max thingy is a bit exploratory right now
and very sensitive to the algorithms parameters. I've updated things in
the last few days (I need this feature for temperature compensation code
which I'm working on), so it should be in better shape.
Vijay, can you send me the models you were dealing with so I can check
things out?
Thanks,
Emery
Herbert Sauro wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Vijay tried to use Oscill8 to determine the max and min for an
oscillatory regime but was unable to for some reason, instead he did it
himself using Matlab. Could you email him to ask him for the models in
question and figure out what was wrong?
Thanks
Herbert
</pre>
<pre wrap=""><hr size="4" width="90%">
<?xml version="1.0" encoding="UTF-8"?>
<!-- Created by XMLPrettyPrinter on 8/5/2005 from JDesigner 2.0.11 Beta -->
<sbml level = "2" version = "1" xmlns = <a
class="moz-txt-link-rfc2396E" href="http://www.sbml.org/sbml/level2">"http://www.sbml.org/sbml/level2"</a> xmlns:jd = <a
class="moz-txt-link-rfc2396E" href="http://www.sys-bio.org/sbml">"http://www.sys-bio.org/sbml"</a>>
<!-- -->
<!-- Model Starts Here -->
<!-- -->
<model id = "untitled">
<annotation>
<jd:header>
<VersionHeader JDesignerVersion = "2.0"/>
<ModelHeader Author = "Mr Unknown" ModelTitle = "No Title" ModelVersion = "0.0"/>
</jd:header>
<jd:display>
<SBMLGraphicsHeader BackGroundColor = "FFEFFFFF"/>
</jd:display>
</annotation>
<listOfCompartments>
<compartment id = "compartment" size = "1">
<annotation>
<jd:display>
<boundingBox h = "0" w = "0" x = "0" y = "0"/>
</jd:display>
</annotation>
</compartment>
</listOfCompartments>
<listOfSpecies>
<species boundaryCondition = "false" compartment = "compartment" id = "x1" initialConcentration = "0">
<annotation>
<jd:display>
<boundingBox h = "32" w = "48" x = "427" y = "62"/>
<pictogram boundarySpeciesStyle = "bsVertVice" complexName0 = "" complexName1 = "" complexName2 = ""
complexName3 = "" edgeColor = "FF969696" edgeThickness = "1" endColor = "FFFFFFFF"
gradientType = "gtLinear" namePosition = "npCenter" pictoComplexType = "cstSimple" pictoShape = "ptRoundSquare"
startColor = "FFFFFFCC"/>
<font fontColor = "0" fontName = "Arial" fontSize = "8" fontStyle = ""/>
</jd:display>
</annotation>
</species>
<species boundaryCondition = "false" compartment = "compartment" id = "y1" initialConcentration = "1">
<annotation>
<jd:display>
<boundingBox h = "32" w = "48" x = "442" y = "118"/>
<pictogram boundarySpeciesStyle = "bsVertVice" complexName0 = "" complexName1 = "" complexName2 = ""
complexName3 = "" edgeColor = "FF969696" edgeThickness = "1" endColor = "FFFFFFFF"
gradientType = "gtLinear" namePosition = "npCenter" pictoComplexType = "cstSimple" pictoShape = "ptRoundSquare"
startColor = "FFFFFFCC"/>
<font fontColor = "0" fontName = "Arial" fontSize = "8" fontStyle = ""/>
</jd:display>
</annotation>
</species>
<species boundaryCondition = "false" compartment = "compartment" id = "b" initialConcentration = "0">
<annotation>
<jd:display>
<boundingBox h = "32" w = "48" x = "513" y = "90"/>
<pictogram boundarySpeciesStyle = "bsVertVice" complexName0 = "" complexName1 = "" complexName2 = ""
complexName3 = "" edgeColor = "FF969696" edgeThickness = "1" endColor = "FFFFFFFF"
gradientType = "gtLinear" namePosition = "npCenter" pictoComplexType = "cstSimple" pictoShape = "ptRoundSquare"
startColor = "FFFFFFCC"/>
<font fontColor = "0" fontName = "Arial" fontSize = "8" fontStyle = ""/>
</jd:display>
</annotation>
</species>
<species boundaryCondition = "true" compartment = "compartment" id = "Node3" initialConcentration = "0.1">
<annotation>
<jd:display>
<boundingBox h = "32" w = "48" x = "291" y = "76"/>
<pictogram boundarySpeciesStyle = "bsVertVice" complexName0 = "" complexName1 = "" complexName2 = ""
complexName3 = "" edgeColor = "FF969696" edgeThickness = "1" endColor = "FFFFFFFF"
gradientType = "gtLinear" namePosition = "npCenter" pictoComplexType = "cstSimple" pictoShape = "ptRoundSquare"
startColor = "FFFFFFCC"/>
<font fontColor = "0" fontName = "Arial" fontSize = "8" fontStyle = ""/>
</jd:display>
</annotation>
</species>
</listOfSpecies>
<listOfReactions>
<reaction id = "J2" reversible = "false">
<annotation>
<jd:display arrowFillColor = "FF648483" lineColor = "FF0099FF" lineThickness = "2">
<lineType type = "ltBezier">
<pt x = "0" y = "0"/>
<pt x = "382" y = "81"/>
<pt x = "382" y = "81"/>
</lineType>
</jd:display>
</annotation>
<listOfReactants>
<speciesReference species = "Node3" stoichiometry = "1"/>
</listOfReactants>
<listOfProducts>
<speciesReference species = "x1" stoichiometry = "1"/>
</listOfProducts>
<kineticLaw>
<math xmlns = <a class="moz-txt-link-rfc2396E"
href="http://www.w3.org/1998/Math/MathML">"http://www.w3.org/1998/Math/MathML"</a>>
<apply>
<minus/>
<apply>
<divide/>
<apply>
<times/>
<ci>
a1
</ci>
<ci>
b
</ci>
</apply>
<apply>
<plus/>
<ci>
k1
</ci>
<apply>
<power/>
<ci>
y1
</ci>
<ci>
n
</ci>
</apply>
</apply>
</apply>
<apply>
<times/>
<ci>
g1
</ci>
<ci>
x1
</ci>
</apply>
</apply>
</math>
<listOfParameters>
<parameter id = "a1" value = "0.2"/>
<parameter id = "k1" value = "1"/>
<parameter id = "n" value = "4"/>
<parameter id = "g1" value = "0.1"/>
</listOfParameters>
</kineticLaw>
</reaction>
<reaction id = "J1" reversible = "false">
<annotation>
<jd:display arrowFillColor = "FF648483" lineColor = "FF0099FF" lineThickness = "2">
<lineType type = "ltBezier">
<pt x = "0" y = "0"/>
<pt x = "390" y = "111"/>
<pt x = "390" y = "111"/>
</lineType>
</jd:display>
</annotation>
<listOfReactants>
<speciesReference species = "Node3" stoichiometry = "1"/>
</listOfReactants>
<listOfProducts>
<speciesReference species = "y1" stoichiometry = "1"/>
</listOfProducts>
<kineticLaw>
<math xmlns = <a class="moz-txt-link-rfc2396E"
href="http://www.w3.org/1998/Math/MathML">"http://www.w3.org/1998/Math/MathML"</a>>
<apply>
<minus/>
<apply>
<divide/>
<ci>
a2
</ci>
<apply>
<plus/>
<ci>
k2
</ci>
<apply>
<power/>
<ci>
x1
</ci>
<ci>
n
</ci>
</apply>
</apply>
</apply>
<apply>
<times/>
<ci>
g2
</ci>
<ci>
y1
</ci>
</apply>
</apply>
</math>
<listOfParameters>
<parameter id = "a2" value = "0.15"/>
<parameter id = "k2" value = "1"/>
<parameter id = "n" value = "4"/>
<parameter id = "g2" value = "0.1"/>
</listOfParameters>
</kineticLaw>
</reaction>
<reaction id = "J15" reversible = "false">
<annotation>
<jd:display arrowFillColor = "FF648483" lineColor = "FF0099FF" lineThickness = "2">
<lineType type = "ltBezier">
<pt x = "0" y = "0"/>
<pt x = "426" y = "97"/>
<pt x = "426" y = "97"/>
</lineType>
</jd:display>
</annotation>
<listOfReactants>
<speciesReference species = "Node3" stoichiometry = "1"/>
</listOfReactants>
<listOfProducts>
<speciesReference species = "b" stoichiometry = "1"/>
</listOfProducts>
<kineticLaw>
<math xmlns = <a class="moz-txt-link-rfc2396E"
href="http://www.w3.org/1998/Math/MathML">"http://www.w3.org/1998/Math/MathML"</a>>
<apply>
<minus/>
<apply>
<times/>
<ci>
c1
</ci>
<ci>
y1
</ci>
</apply>
<apply>
<times/>
<ci>
g
</ci>
<ci>
b
</ci>
</apply>
</apply>
</math>
<listOfParameters>
<parameter id = "c1" value = "0.01"/>
<parameter id = "g" value = "0.01"/>
</listOfParameters>
</kineticLaw>
</reaction>
</listOfReactions>
</model>
</sbml>
</pre>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</body>
</html>
|
|
From: Emery C. <ec...@vt...> - 2005-06-07 16:38:47
|
That would be best for me, naturally! ODE is a general enough format to be useful anyway (a user can then use your SBML=>ODE converter and use XPPAUT directly if they like). Thanks. I will try to get this SBW and stability stuff sorted out in the next two weeks. Emery Herbert Sauro wrote: >Thanks Emry for the email. I realise now that 1.9.7 is not for release just testing. I think we could help you out with a SBML ode converter. I presume you want the format that oscili8 accepts (which is ok by me)? > >Herbert > >-----Original Message----- >From: Emery Conrad [mailto:ec...@vt...] >Sent: Tue 6/7/2005 5:52 AM >To: osc...@li... >Cc: >Subject: Re: [Oscill8-devel] Re: inspector >Herb, > >I agree. Technically, the bsam branch isn't for release; this is my testing branch for the new features I'm currently developing. It wasn't actually intended for anyone to use, but a few folks asked me to release it on sourceforge so I did. I'll release it officially later in June and it should be stable and working at that time. > >Also, I'm going to switch the installation over to a USER installation, rather than an ADMIN install. This will make things less dependent on security; there are students who want to use Oscill8, but can't because they don't have admin privileges on their machines. I will change this for the next "real" release. > >About it's fragility for model input, that's for sure. If you get the model just right, everything works smoothly (most of the time). But that step isn't yet made simple for the user. I plan on making that a bit more straight-forward in the future. > >If you guys have spare time and would like to, an SBW module that converts SBML --> ODE would be nice... You basically have that in Jarnac already, except you don't remove differential eqs for dependent species. All you'd need to do would be to remove ODEs for those dependent species, and then you'd have a bifurcation ready model. > >Anyway, I'm in the process of adding a few fundamental new features right now; the bsam branch is going to remain rather unstable for another week or two at least. At that point, I would really like to finalize the SBW connection so that we can all use things smoothly through JDesigner etc. If you have the time and/or the desire, adding a Jarnace getReducedODEList() would be nice... that would do it for bifurcation analysis. > >Take it easy, > >Emery > > >Herbert Sauro wrote: > >Emery, version 1.9.7 is not well, it's quite a difficult version to get > >running. The dependence on jigcell is one issue, I don't think many > >users would realize that they must install jigcell. Also it's quite > >picky about models, some work some don't (I think it's the jigcell bit > >which is having problems). Also on machines which are connected to the > >network there are security issues with using program files to sort > >temporary information, i.e it won't let your software save files to this > >area (I have the same problem with my software). > > > >If there is anything we can do let me know. > > > >Herbert > > > >-----Original Message----- > >From: Emery Conrad [ mailto:ec...@vt...] > >Sent: Monday, June 06, 2005 11:15 AM > >To: osc...@li... > >Subject: [Oscill8-devel] Re: inspector > > > >Herb, > > > >1.9.7 is on the bsam branch, which is unstable at this point; the HUH? > >error happens when Oscill8 doesn't know where the install directory > >is... Frank said he was having this trouble too... it might work if you > >reboot (did you install Oscill8 to a different directory than previous > >installs... this might cause it as well). Otherwise, make a c:/log > >directory; that should be a temporary fudge. > > > >This SBW interface relies on JigCell right now, which is broken for sbml > >conversion. Best way for now is to get the ODEs from JDesigner or > >something, and remove the conservation relations by hand, defining extra > >"total" parameters as needed. > > > >Sorry I couldn't be more helpful... after next week Monday I'll put > >effort into getting things running smoothly with SBW. > > > >Emery > > > > > > > >Herbert Sauro wrote: > > > > > >Frank just showed me the lastest source under the bsam branch. > > > >However, I still get a crash when I call oscill8 from sbw. > > > >Herbert > > > > > > > > > > > > > > > >------------------------------------------------------- > >This SF.Net email is sponsored by: NEC IT Guy Games. How far can you > >shotput a projector? How fast can you ride your desk chair down the > >office luge track? > >If you want to score the big prize, get to know the little guy. > >Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20 > >_______________________________________________ > >Oscill8-devel mailing list > >Osc...@li... > >https://lists.sourceforge.net/lists/listinfo/oscill8-devel > > > > > >------------------------------------------------------- This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput a projector? How fast can you ride your desk chair down the office luge track? If you want to score the big prize, get to know the little guy. Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20 _______________________________________________ Oscill8-devel mailing list Osc...@li... https://lists.sourceforge.net/lists/listinfo/oscill8-devel > > > > > |
|
From: Emery C. <ec...@vt...> - 2005-06-07 12:52:38
|
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
Herb,<br>
<br>
I agree. Technically, the bsam branch isn't for release; this is my
testing branch for the new features I'm currently developing. It wasn't
actually intended for anyone to use, but a few folks asked me to
release it on sourceforge so I did. I'll release it officially later in
June and it should be stable and working at that time.<br>
<br>
Also, I'm going to switch the installation over to a USER installation,
rather than an ADMIN install. This will make things less dependent on
security; there are students who want to use Oscill8, but can't because
they don't have admin privileges on their machines. I will change this
for the next "real" release.<br>
<br>
About it's fragility for model input, that's for sure. If you get the
model just right, everything works smoothly (most of the time). But
that step isn't yet made simple for the user. I plan on making that a
bit more straight-forward in the future.<br>
<br>
If you guys have spare time and would like to, an SBW module that
converts SBML --> ODE would be nice... You basically have that in
Jarnac already, except you don't remove differential eqs for dependent
species. All you'd need to do would be to remove ODEs for those
dependent species, and then you'd have a bifurcation ready model.<br>
<br>
Anyway, I'm in the process of adding a few fundamental new features
right now; the bsam branch is going to remain rather unstable for
another week or two at least. At that point, I would really like to
finalize the SBW connection so that we can all use things smoothly
through JDesigner etc. If you have the time and/or the desire, adding a
Jarnace getReducedODEList() would be nice... that would do it for
bifurcation analysis.<br>
<br>
Take it easy,<br>
<br>
Emery<br>
<br>
<br>
Herbert Sauro wrote:
<blockquote
cite="mid...@DE..."
type="cite">
<pre wrap="">Emery, version 1.9.7 is not well, it's quite a difficult version to get
running. The dependence on jigcell is one issue, I don't think many
users would realize that they must install jigcell. Also it's quite
picky about models, some work some don't (I think it's the jigcell bit
which is having problems). Also on machines which are connected to the
network there are security issues with using program files to sort
temporary information, i.e it won't let your software save files to this
area (I have the same problem with my software).
If there is anything we can do let me know.
Herbert
-----Original Message-----
From: Emery Conrad [<a class="moz-txt-link-freetext" href="mailto:ec...@vt...">mailto:ec...@vt...</a>]
Sent: Monday, June 06, 2005 11:15 AM
To: <a class="moz-txt-link-abbreviated" href="mailto:osc...@li...">osc...@li...</a>
Subject: [Oscill8-devel] Re: inspector
Herb,
1.9.7 is on the bsam branch, which is unstable at this point; the HUH?
error happens when Oscill8 doesn't know where the install directory
is... Frank said he was having this trouble too... it might work if you
reboot (did you install Oscill8 to a different directory than previous
installs... this might cause it as well). Otherwise, make a c:/log
directory; that should be a temporary fudge.
This SBW interface relies on JigCell right now, which is broken for sbml
conversion. Best way for now is to get the ODEs from JDesigner or
something, and remove the conservation relations by hand, defining extra
"total" parameters as needed.
Sorry I couldn't be more helpful... after next week Monday I'll put
effort into getting things running smoothly with SBW.
Emery
Herbert Sauro wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Frank just showed me the lastest source under the bsam branch.
However, I still get a crash when I call oscill8 from sbw.
Herbert
</pre>
</blockquote>
<pre wrap=""><!---->
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you
shotput a projector? How fast can you ride your desk chair down the
office luge track?
If you want to score the big prize, get to know the little guy.
Play to win an NEC 61" plasma display: <a class="moz-txt-link-freetext" href="http://www.necitguy.com/?r=20">http://www.necitguy.com/?r=20</a>
_______________________________________________
Oscill8-devel mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Osc...@li...">Osc...@li...</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/oscill8-devel">https://lists.sourceforge.net/lists/listinfo/oscill8-devel</a>
</pre>
</blockquote>
</body>
</html>
|
|
From: Emery C. <ec...@vt...> - 2005-06-06 19:15:28
|
Herb, 1.9.7 is on the bsam branch, which is unstable at this point; the HUH? error happens when Oscill8 doesn't know where the install directory is... Frank said he was having this trouble too... it might work if you reboot (did you install Oscill8 to a different directory than previous installs... this might cause it as well). Otherwise, make a c:/log directory; that should be a temporary fudge. This SBW interface relies on JigCell right now, which is broken for sbml conversion. Best way for now is to get the ODEs from JDesigner or something, and remove the conservation relations by hand, defining extra "total" parameters as needed. Sorry I couldn't be more helpful... after next week Monday I'll put effort into getting things running smoothly with SBW. Emery Herbert Sauro wrote: > >Frank just showed me the lastest source under the bsam branch. > >However, I still get a crash when I call oscill8 from sbw. > >Herbert > > > |
|
From: Emery C. <ec...@vt...> - 2005-05-09 19:02:11
|
Herb and Frank, I've given you developer privileges to sourceforge Oscill8 (repo is at /cvsroot/oscill8); I also created this developer list to send messages specific to development. I've removing and changing a lot right now, so if you decide you feel like doing something to the code in the near term, it's probably best to email this list (osc...@li...) to talk about it. Naturally, feel free to ask me to add new features if you have any good ideas. Eventually, I'd probably like to ask Frank for a bit of advice for better graphics behavior; maybe when you're out here, Frank. I did the first, simple approach which probably wasn't the best. BTW, I've just updated Oscill8 to do SBML-->ODE conversion and to run from the command line directly on an ODE or SBML file (it creates a workspace in the Oscill8/wscache directory); this is in version 1.1.1. Take it easy, Emery |
