Nabble - Simbrain

Re: Noise generator. How it working?

2009-08-13T22:21:06Z

Thank you for your answer, Jeff.

Re: Noise generator. How it working?

2009-08-12T15:05:51Z

Hi Alexander,

Noise is added simply by creating a random number within specified bounds, according to a user-specified rule, and then adding that number to current activation (how the random number is actually used depends on the neuron / synapse type). The source code is pretty straightforward; you can browse it here:

http://code.google.com/p/simbrain/source/browse/trunk/src/org/simbrain/network/util/RandomSource.java

We didn't consult any theoretical neuroscience texts when we created this class. I know the topic of noise in neural networks and computational biology is a big topic, but it's not one I'm any expert in (anyone else want to chime in?). I just did a few quick google scholar searches and "neural noise" yielded quite a few hits.

If memory serves, Poisson distributions are often used to model noise in neural networks. We have not implemented this, thought the RandomSource class is pretty flexible, and could easily be extended to incorporate more types of noise.

As usual, suggestions are welcome.

Best,

- Jeff

Noise generator. How it working?

2009-08-11T03:08:42Z

Hello, Jeff!
Can you tell me some words about noise generator in SimBrain...
1. Noise generator inject current of random size to neuron in each step of simulation. Am I right?
2. Can you give me а link to the article (or attach article) in wich this approach is discribed. I want to read some more about noise generation in biological neural networks, but I don't know where.

Thank you.
Alexander.

Re: Getting started..

2009-08-10T07:50:03Z

1.6 and Vista should both be fine. Did it work when you double clicked on the jar? What happened?

- Jeff

Re: Getting started..

2009-08-09T18:38:00Z

i have java 1.6 installed, since i am a student.does this work in vista?

Re: Getting started..

2009-08-09T18:06:16Z

It's a java desktop application, packaged as a jar file. There is no ".exe" file. You should be able to just double click the Simbrain.jar file, which is in the directory you unzipped. If you do that, and nothing happens, then you probably have to install java (1.4+) on your system. You just need the JRE (java runtime environment), which you can get here:

http://www.java.com/en/download/manual.jsp

Cheers,

- Jeff

Getting started..

2009-08-08T02:06:57Z

i downloaded the SimBrain zip file and unpacked it.i can't find any executable.the reply given in the previous topic doesn't seem to work.kindly help

Re: Start up help request

2009-05-15T02:27:14Z

Hi Jeff,

Many thanks for the feedback. Still having a few issues, but I'll take them offline with you, and return to your question about how BP might best be presented in Simbrain.

Setting up the network at present is simple, so no change needed there.

Data management can be tricky, but the basics are already just about available. I would suggest that rather than loading training data and output data through the subnetwork menu, instead the user ought to be encouraged to open up data worlds: one each for training, testing, and validation, if that is done. It would be good to be able to keep the input patterns and the desired targets in the same csv file.

I wouldn't work too hard on facilitating automated training, error monitoring, verification and testing. Instead, the user should run the network with a single data world at a time. Manually changing the I/O dataset to perform verification by hand, and to test at the end, is not too onerous, and while not what one would want in an optimized system, might be a good idea for a teaching system. No special suport for verification and testing are necessary, as long as the user remembers to clamp the weights before using them.

However a key utility that is missing and that would make all this work well in a teaching setting is graphing: continuous graphing of error is essential to make sense of the training protocol. I know you are working on some graphing for 3.0, but I would suggest it would be a good thing to prioritize. Plotting the individual activation functions in input space is also useful, as for example here: http://cspeech.ucd.ie/~fred/teaching/oldcourses/ann98/bpexample.html

I'd be happy to discuss further, or help out, of course!

Fred

Re: Start up help request

2009-05-14T10:41:34Z

Hi Fred,

Backprop is not one of Simbrain's best features, currently. The user interface is a bit counter-intuitive (as you've discovered), and it has not received as much attention as other aspects of the program. Part of the reason for this is that backprop is well supported in other open source packages (e.g. emergent, joone). It is one of the many things being improved for version 3.

Having said that, it does work, the testing mechanism you describe does exist, and XOR can certainly be trained in less than a million iterations! The docs on this are here, but they are not as clear as they could be.

Here's what you have to do.

First, create a new backprop network (Insert > New Network > Backprop) with 2 input nodes, 1 output node, and as many hidden units as you want. It sounds like you were able to do this.

Second, right click on the "backprop" tab, and select "Train." All the action will take place inside this dialog.

Third, inside the dialog, click the randomize button.

Fourth, click the input file and output file buttons, and attach data files to them. For input, use Simbrain > simulations > networks > bp > training > xor_in.csv, and similarly use xor_out.csv for output. I think you made it this far.

Fifth, click the play button. If everything is set up right, you will see a representation of the current error, and you can just run it until it reaches a level you are happy with. There is also a "batch" tab where you can run for a set number of iterations. Since this bypasses the Gui it's faster. (One problem here is that if you set up the input and output files wrong, you don't get any error message.)

When you're done, press Ok. You can then test the network by selecting the input nodes and pressing the up and down arrow keys on your keyboard to set various input patterns, and iterating the network.

If I should clarify any of this let me know!

Let me end by asking you (and others) a question. One thing I've thought a lot about, but have never decided on, is what the most intuitive way would be to implement backprop in a Gui. I want to do it in a way that is consistent with Simbrain's emphasis on visual intuition and ease of use, but as a user interface problem backprop is a bit challening. The problem is there are all these steps that require different types of configuration. Setting up the network, setting up training and testing data, training the network, cross-validating, pre-processing data, etc. I have a rough idea of how I want to do these things in the next version of Simbrain, but I'd be curious what others think.

Best,

- Jeff

Re: Start up help request

2009-05-14T04:19:14Z

Update: I have discovered that one can create a BP-network as a specific type of subnetwork, and then there is a training option that requests input and output files.

Training seems wrong to me though. On a simple 2-2-1 network, training on XOR, with a learning rate of 0.9 and some momentum, the network is taking ages (literally millions) of iterations to get anywhere, and it still isn't solving XOR. The error appears briefly on screen, but doesn't seem to be logged anywhere. I have trained networks with these parameters many times, and the solution should be reached quickly.

To test the network, I open the input patterns as a data world, and step through them by clicking on each pattern in turn. But this is again not logged and also cumbersome. Is there a formal testing mechanism that will cycle through the input patterns and produce some quantitative overview of how the network is performing?

All help welcome.

Fred

Start up help request

2009-05-14T04:04:25Z

Hi,

I find the range of models covered in Simbrain to be quite exciting, and I am sure 3.0 will be even richer. Oddly, however, I find it difficult to do what I (perhaps wrongly) would consider to be the more basic ANN operations. Perhaps these are not covered, but more likely, I am just not finding the right means.

I want to start with a simple BP network, such as the XOR network here. I would like to train the network, and observe the decrease in error.

I can open, or create the network, just fine. I can also open two data-world sets, one for input patterns and one for output patterns. I change the 'clamped' default of the input units to be 'linear' and set them up so that each receives input from one column of the input data table. I set the data table to be in 'iterative' mode. Now when I press Run, the network cycles through the patterns.

But I am not using the output patterns, nor do I see any commands to train the network or monitor its output error. I know that there is a BP lesson where the network has a box around it, a BackPropagation label, and if I click on that I get the option to train the network, but I don't see it for a network I create myself, of for the XOR network.

Any help greatly apprecitated!

Fred

Delighted to find Simbrain

2009-05-13T07:01:13Z

Having just found Simbrain, I simply want to express my delight at finding a well implemented simulator that looks to be ideal for pedagogical purposes. I teach a short course in connectionism to masters students in cognitive science, many of whom have esentially no computer background. We have made do with tLearn and JNNS over the last few years, but with lots of problems. Simbrain looks perfectly pitched for this purpose, and I look forward to using it! Many thanks!

Simbrain 2.01 Released

2009-03-04T13:23:23Z

All:

In response to recent threads I have uploaded a new version of Simbrain 2, which changes "seconds" to "msec", adds better documentation and defaults for Izhikevich neurons, and some minor feature enhancements for the datatable.

http://code.google.com/p/simbrain/downloads/list

If anyone catches bugs let me know; this is the first time an iterative release has been created.

Of course, Simbrain 3 is still the main focus of development efforts now...

- Jeff

Re: Izhikevich neuron has no stable state

2009-02-15T22:52:22Z

Thanks very much. Now I understand. I'll try to construnct my net again as soon as possible.

Re: Izhikevich neuron has no stable state

2009-02-13T20:48:14Z

Hi Casper,

You are absolutely right. It's a little embarrassing that I've missed this for so long. I had a look at the Izhikevich article I used to implement the equation, and the Wilson book I used for Naka Rushton, and both differential equations have time in milliseconds. Come to think of it, Neuron and Genesis default to milliseconds, and it seems to be a standard unit in computational biology.

So, the text in the Simbrain GUI should read "msec" instead of "seconds," since that's already what's assumed in all the Simbrain neuron models that are based on differential equations. The mistake is just in that bit of GUI text. Since all these models assume msec, you should be able to mix Naka Rushton and Izhikevich neurons without a problem.

I think this warrants an update to 2.0. I'll try to fix a few other things too (like the Izhikevich default values), and perhaps even scrollbars in the data table. I'll get to it as soon as possible and put the updated "2.01" download on the website.

Thanks very much for raising these issues.

- Jeff

Re: Feature requests for 3.0?

2009-02-13T20:24:36Z

Hi Casper,

These are great suggestions. We have the scroll bars in 3.0 but not the option to just specify a table size using fields, which I'll put on the todo list. It also strikes me that it would be nice to make the table just grow as data are added, like in the case you had from the previous email. I'm not sure what the best interface for doing that is, but I'll think about it and I'm open to suggestions.

Thanks,

- Jeff

Re: Feature requests for 3.0?

2009-02-13T04:00:22Z

Hello!

Small disadvantages of DataWorld.
I worked with many rows, so:
1. I should add rows one by one, or form csv file. In v.3 it will be good to set, how many rows need to be added. For example: "1500 rows, please" :))
2. There is no scroll bar in dataWorld, when opening long CSV. So, i can see only first rows, and others only after closing network and dataWorld in CSV.

Re: Izhikevich neuron has no stable state

2009-02-13T03:38:23Z

Hello, Jeff.

May be i'm not right. But i don't Understand one thing (the same i tried to explain in my previous post). I'll try to explain again, form other side.
Steps, i made:
1. I get Network form your's previous post.
2. Set time-step to 1 (second).
3. Connect output of Izh neuron to DataWorld. Add many Rows to it in CSV file.
4. Run simulation and got а row of states of Izh neuron. Each row is state in each second (with this time-step). Am I right? (see fig.1)

5. I import this data to MatLab and draw plot. (see fig.2)

6. As we can see from fig.2, destination between 1st two spikes is 28 seconds (you see - seconds!!! ). It is impossible!
7. Now, let's take a look to fig.3, taken from one of Izhikevich's articles.

You see, destination between pulses is about 20-40 ms (!! milliseconds) (look at small line, wich is eqal to 20 ms at right-bottom coner).
8. So, as I can understand, model is working propelly, but timeframe is wrong? if we consider, that time in SimBrain for this model not in seconds, but in milliseconds, they will be eqal.
(
9. If i set time-step to .001. In data world each line will be one millisecond. So, after 100-150 iteretions, i hope get smth like fig.3, but pulses are overblowed.
)
==
May be, I don't understand anything again?
==
It's not a problem, if i'll use only Izh neurons. But i want to combine them with Naka-Rushton model. and i can't, because of various timeframes.
==

Thank you.

Re: Izhikevich neuron has no stable state

2009-02-12T10:09:18Z

Hello again,

A friend and I did some work validating Simbrain 2's Izhikevich implementation against the matlab scripts, including the tonic spiking case. We did not find any discrepancies, so perhaps you can send more detail about the issue you ran into?

One thing that was an issue is that the time-step in Simbrain does not, by default, match the tau parameter in the matlab scripts. So that should be set to .2

Attached is a network that simulates tonic spiking with input current of 14.

Let me know if you have more questions or run into problems.

- Jeff

TonicSpiking_A.net

Re: Izhikevich neuron has no stable state

2009-01-28T22:57:58Z

Hello,

Though I said this was somewhat complex, it actually isn't (things are more complex for the "post-synaptic" neuron of a spiking neuron, which must convert a spike into a number, but that's not at issue here). The input current to the Izhikevich neuron is just weighted input. So, assuming the weight is set at its default value of 1, then yes, the activation of the clamped neuron in your example just is the current I.

So, if you are getting different results in Simbrain than you do with the matlab script, then it may be something with the current implementation of the Izhikevich model (sigh...).

There are a few things I can think of that might be causing the discrepancy, but I have not looked at it myself and alas, my next few days are pretty busy. I'll try to have a look this weekend. If you're itching to figure this out before then email me or post another message and I can at least tell you what my suspicions are.

- Jeff

Re: Feature requests for 3.0?

2009-01-28T22:44:48Z

Hello again,

I've been thinking about your interesting post all day. I like your idea of a two-stage approach. But I want to try to make the first stage so usable that a second stage won't be necessary. If I fail, then of course the "equation editor model" you describe can always be added.

Before I say more, a digression. I think of "ease-of-use" as something that should extend to the innards of a piece of software. It's so nice to open an xml file and be able to understand its contents, or follow a "literate" program. One thing I've really been trying to achieve with the next version of Simbrain (and with lots of help, and on the basis of suggestions by others) is a clean API, so that users can use parts of Simbrain as a library in their own programs. (As of v3 Simbrain will be released in several forms, including a small neural network jar). So, the hope is that one person could use Simbrain as a GUI, another could use it as a library in their own program , another person could add their own neuron and synapse types, another could build little applet demos using the libraries, etc. Think of Mac OS X. It's this nice fluffy GUI, but there is also a terminal you can open which power users can take advantage of.

Ok, back to the point. The way I'm thinking of it is that you'd open the "custom neuron" dialog, and there would be a little little editor pane (with a help button on top which would take you to a document explaining how to do simple things). Users would not even know it was beanshell. The text you would enter might look something like this.

a = 10;
b = 20;
double val = getInputValue() + getActivation() + a + Math.sin(b);
setBuffer(clip(val)); // clip ensures that the value is between upper and lower bounds

I think a "power user" (e.g. someone who can understand the Hodgkin Huxley equations) could master this pretty easily, after seeing a few examples and experimenting, and use it to prototype their own custom activation functions. Something similar could be done for weights and subnetworks.

At any rate, with this structure in place, it might not be too hard to move to the more graphical "equation editor" model, using something like the method you describe in your "Step 2."

Thanks for the thought provoking suggestions,

- Jeff

Re: Izhikevich neuron has no stable state

2009-01-28T04:41:50Z

It's me again. You say:

jyoshimi wrote:

It should be fun to try these out (just open the neuron property dialog for an Izhikevich neuron and enter appropriate values. For I values it's a little more complicated and I can explain if you ask; but most of them are 0 anyway).

I try all you suggested me. It's working! But i have a qestion again. :( About net, You'v gave me. It has 2 neurons. 1st - clamped, 2nd - Izhikevich. I set activation of 1st neuron to 40, 2nd neuron fireing. Great! But, what is "Activation" (40) for the 1st neuron? Is it current to 2nd neuron (40mV)? I could not understand, so:

I made another experiment. I set params of 2nd as for A neuron (tonic spiking). Connect to it Clamped neuron with Activation = 14 (if it is current, should it fire?), but nothing has happend...

I download *.m file from site, you'v advice me. And test Izh neuron with current = 14mV, and it fires as in picture. So, activation is not a current! Am I right?

As we can see from picture, tonic spiking neuron should fire each 50 ms. To achive this result, activation of 1st neuron should be 280. Why?

May be i don't understand something again? Could you explain me about "I" values? May be this is the problem?

Re: Feature requests for 3.0?

2009-01-28T00:46:22Z

Yes, graph is great! No other words!

About eqatations... I read your article at
http://www.brains-minds-media.org/archive/1411
and, as i see from it, you are using your application in studing prosess in social studies. Am i write? I think it would be hard to explain to students java programming. So, it should be something simple. May be, something like Microsoft Equation.... Student select signs and operations from panel and fill its params.. But i think, it is hard to realize.

As for me (and, i think, many others) BeansShell and link to the langs syntax will be enough, but... One of project's goals is "Easy-to-use, visual GUI ... ets.". So, is it sutable?

May be this approach:
Step 1. Inserting BeansShell scripting in one version (may be 3.1) (small efforts, as i understand).
Step 2. In next (or next, next, ..., if the interest would be) insert any GUI, may be, like i proposed. This GUI will have button "Get script" wich will convert operations from panel into BeansShell script. (Support of scrips was realized in first step.) User could see this script and correct it, if need.

Re: Izhikevich neuron has no stable state

2009-01-27T23:50:57Z

Thank you very much for the answer and link! It is very interesting. Yesterday I read another article: http://vesicle.nsi.edu/users/izhikevich/publications/spikes.pdf
and simulate matlab algorithm. I think, that the best is picture 2 of fig.2. Its good for understanding the meaning of parameters. Tonight i'll try your net and matlab script from site.

jyoshimi wrote:

As I noted in another post on this forum just now, I only have time to work on this in my spare time so not everything works perfectly...

No, no! Everything is great! That problem was because i do't understand the model and its params propely and also i understan how hard it is to write such application in spare time!

And, if it will interest you, here is another simulator (as toolbox for matlab):
http://www.ymer.org/amir/software/biological-neural-networks-toolbox/

Thanks a lot!

Simbrain Development now at Google Code

2009-01-27T21:04:43Z

Just a note that Simbrain development has been moved over to google code:

http://code.google.com/p/simbrain/

- Jeff

Re: Izhikevich neuron has no stable state

2009-01-27T21:01:45Z

The Izhikevich model can capture lots of different types of neuron behavior, including behavior where it simply bursts and spikes spontaneously (like real neurons do). Check out this site:

http://vesicle.nsi.edu/users/izhikevich/publications/whichmod.htm#izhikevich

If you have matlab then the script on that site (which I just discovered today) is a nice way to get a feel for the model. I've attached a small network file (for Simbrain v. 2; see the bottom of this post) that has the parameter values for the "integrator," which I am guessing is something like the integrate and fire model. It stays in a stable state absent input but it increases its rate of firing as the value of the neuron on the left is increased (just select it and press the up button).

The matlab script at Izhikevich's site has a nice summary of parameter values for different types of neural behaviors:

A B C D I (Input from other neurons)
0.02 0.2 -65 6 14 % tonic spiking
0.02 0.25 -65 6 0.5 % phasic spiking
0.02 0.2 -50 2 15 % tonic bursting
0.02 0.25 -55 0.05 0.6 % phasic bursting
0.02 0.2 -55 4 10 % mixed mode
0.01 0.2 -65 8 30 % spike frequency adaptation
0.02 -0.1 -55 6 0 % Class 1
0.2 0.26 -65 0 0 % Class 2
0.02 0.2 -65 6 7 % spike latency
0.05 0.26 -60 0 0 % subthreshold oscillations
0.1 0.26 -60 -1 0 % resonator
0.02 -0.1 -55 6 0 % integrator
0.03 0.25 -60 4 0 % rebound spike
0.03 0.25 -52 0 0 % rebound burst
0.03 0.25 -60 4 0 % threshold variability
1 1.5 -60 0 -65 % bistability
1 0.2 -60 -21 0 % DAP
0.02 1 -55 4 0 % accomodation
-0.02 -1 -60 8 80 % inhibition-induced spiking
-0.026 -1 -45 0 80 % inhibition-induced bursting

It should be fun to try these out (just open the neuron property dialog for an Izhikevich neuron and enter appropriate values. For I values it's a little more complicated and I can explain if you ask; but most of them are 0 anyway). I'll include this list in the next version of the documentation.

Now, since you asked about this, I'll point out that I implemented the Izhikevich neuron quickly and without any optimization, and I never really tested it. So I'd love it if you or anyone in the open source community could validate (or even optimize!) the behavior of Simbrain's implementation of the Izhikevich model. As I noted in another post on this forum just now, I only have time to work on this in my spare time so not everything works perfectly...

All best,

- Jeff

izhik_test.net

Re: Feature requests for 3.0?

2009-01-27T20:35:36Z

Well, the hope is to have it done some time this year, but it's hard to say. At the very least I'm sure we can get an "alpha" release out. Most of the major refactoring is done (it took a long, long time to get certain things right). But now there is a ton of finish work to do. This isn't my main work so I can only work on it in my spare time. So, if anyone wants to help out! =)

Also note that we have moved development over to google code,

http://code.google.com/p/simbrain/

which I'm pretty excited about. I'll announce that in a separate thread.

- Jeff

Re: Feature requests for 3.0?

2009-01-27T20:31:06Z

Hi Alexander (or Casper?),

Thanks for your enthusiasm!

"Do u plan to insert GRAPH module to display neuron activity at timeline?"

Yes, we are using JFreeChart for v3, which has lots of great charting capabilities. Here is a screenshot with the kind of time-series you're asking about.

"And other suggession: can you insert field so i can write there my own eqatations? For example Hoghin-Hagsley neuron eqatations. (I want to test my own models). Is it hard to do?"

Others have requested this in the past, and it's been on the todo list for a long time, but because you're asking for it I'll bump it up in priority! One question is what the best user experience would be for this. What would be the most intuitive way to implement this? I'd be interested in what you and others think.

But realistically, I think the easiest would be to give users the ability to create custom activation functions using a small beanshell script (we have expanded beanshell scripting abilities for version 3). This would mean writing a few lines of java code to implement your custom activation function (the same could be done for synapses and subnetworks too).

But again, I am curious how you and others would intuitively want to see this implemented.

Best,

- Jeff

Izhikevich neuron has no stable state

2009-01-27T00:58:29Z

Hello Jeff, I have a small qestion.

First, I make new network. Then I put Izikevich neuron to workplace and push "simulate" button. I see, that it is fireing again and again whithout any affection on it. Can i set it to some stable state? (like integrate-and-fire have it at 0.4?). I run example "SinusoidalInput" (i need to do something like that), but insted of IAF i need to use Izhikevich, but i can't. So, that is the qwestion... Can i do it and how i can? (Sorry for my bad English).

Thanks before.

Re: Feature requests for 3.0?

2009-01-26T23:06:42Z

It's me again! :) When do u plan to finish your 3.0 version?

Re: Feature requests for 3.0?

2009-01-26T23:04:57Z

Hi, this is super programm. I like it! I can say more: I love it! I'm interested in modelling human's vizual analyzer. Do u plan to insert GRAPH module to display neuron activity at timeline? I think it is not very hard? And other suggession: can you insert field so i can write there my own eqatations? For example Hoghin-Hagsley neuron eqatations. (I want to test my own models). Is it hard to do? (Sorry for my bad english)

Feature requests for 3.0?

2008-12-30T15:02:40Z

Hi everyone,

Major refactoring is winding down for version 3. There are a quite a few new features in v. 3, including new components (vision, sound, and a full suite of charts using JFreeChart), scripting ability, improved saving and opening of simulations, a separate neural network jar for use outside of the Gui, and lots more.

But there is still a ton of work to do before the beta release, and so there is time to incorporate new features.

I'd be interested if anyone has any suggestions, in terms of features they'd like to see, lessons they'd like to be able to teach using the software, ideas regarding the GUI design and usability, etc. Even if the suggestion does not make it into the next release, it could make it into a subsequent release, and the feedback would be useful to me.

Best, and happy new year,

- Jeff

Feature requests for 3.0?

2008-12-30T15:02:40Z

Re: Can develop java program with Simbrain?

2008-06-26T11:23:06Z

Nervestaple:

I'm glad for yours (and Hui's) interest in the underlying model networks in Simbrain, especially since the code wasn't originally written to be used as a library. On the other hand, we did really try to make it clean, and separate from the GUI view, and it went through several refactors. I'm especially glad you are happy with the spiking neuron stuff, since I didn't have much guidance when I implemented that (e.g. I had to come up with weird class names like "spikeresponder").

The first thing I'll do is mention SNNAP (http://snnap.uth.tmc.edu/), but you say you've been looking around so you probably saw it already.

In terms of using Simbrain outside of the GUI, this is becoming one emphasis with version 3. So, I'd really like to work with you in implementing your spiking networks outside of the GUI. Perhaps we can talk about what you'd like to accomplish, and then we can spend a few weeks making as much progress as we can. In the process I'll put up some docs, add relevant services in the API, etc. We also plan to improve the updating so it can run efficiently on multi-core machines. Then after a few weeks you can make the call about whether it looks like it will work for your needs. Either way it will push me to do some things on my todo list!

If we do pursue this, it would be great if you jumped in to the code at some point, e.g. to improve the numerical integration implementation (we currently use Euler's method).

Feel free to contact me directly or continue this thread here. There is also a developer list, but since we just switched back to sourceforge it may be migrated there. If it is we could talk there too.

- Jeff

Re: Can develop java program with Simbrain?

2008-06-26T00:59:53Z

Thanks Jeff for saving me so much time!!

Yang, Hui