Re: [dadi-user] Instantaneous population growth/contraction

[Gutenkunst, Ryan N - (rgutenk)]

Hello Ray,

I don't hink I entirely understand your question, so please correct me if this doesn't help.

If you want to model a bottleneck that recovers via a step-function, you would have three Integration.xxx_pop calls: one before the bottleneck, one during the bottleneck, and one after recovery. During each call you can use a different parameter for the population size.

A bottleneck with zero duration would have no effect on the SFS, because no drift would happen during that period.

Best,

Ryan

On 1/29/13 6:43 PM, "tinga...@gmail.com" <tinga...@gmail.com> wrote:

Hi

I'm a new dadi user interested in modeling a demography involving numerous population contractions and expansions. I would like to make the expansion phase following a contraction to be instantaneous rather than exponential. Based on what I have read in the manual and on this forum I am guessing that this could be achieved by setting the time of the bottleneck and the expansion to be the same, but allowing the population sizes in each event to be separately inferred. Would this work? Also how does one enforce explicit expansion or contraction in the Integration.xxx_pop function? 

Regards

Ray

-- 

Ryan Gutenkunst

Assistant Professor

Molecular and Cellular Biology

University of Arizona

phone: (520) 626-0569

http://gutengroup.mcb.arizona.edu

[tinga...@gmail.com]

Thanks for the reply. 

If I understand the following line in your reply correctly:

"A bottleneck with zero duration would have no effect on the SFS, because no drift would happen during that period."

Then there is no loss of genetic variation via the sampling effects induced by the bottleneck, rather any loss is due increased drift whilst at a smaller population size. Is this the case?

For my model I want to simulate instantaneous loss of genetic variation caused by a strong population decrease, then expand this population immediately, so that the newly expanded population contains the same genetic variation as the bottlenecked population (i.e. no new mutations during the bottleneck period). 

If my interpretation of your reply is correct, then perhaps I could achieve something similar to this by creating a parameter, t, for the length of the bottleneck, which is a fixed function of the length of the phylogeny. If t is fixed to be relatively short, then the number of new mutations occurring over the bottleneck period would be kept to a minimum. Would this work, and is there some minimum increment of time that I could use for t? Alternatively, is it possible to set the mutation probability to 0 during certain points of the simulation?

Regards

Ray

[Gutenkunst, Ryan N - (rgutenk)]

Hello Ray,

If you don't want any mutations to be introduced during a specific time period, you can set theta0 = 0 during the corresponding integration period.

On 2/5/13 1:36 PM, "tinga...@gmail.com" <tinga...@gmail.com> wrote:

If I understand the following line in your reply correctly:

"A bottleneck with zero duration would have no effect on the SFS, because no drift would happen during that period."

Then there is no loss of genetic variation via the sampling effects induced by the bottleneck, rather any loss is due increased drift whilst at a smaller population size. Is this the case?

Correct. In the diffusion limit dadi adopts, there is an implicit assumption that the population size is large enough to treat allele frequencies as continuous. In practice, this means the model breaks down if the number of individuals in the population is very small. In practice, "very small" is probably 10s of individuals, but that's just a very rough rule of thumb.

For my model I want to simulate instantaneous loss of genetic variation caused by a strong population decrease, then expand this population immediately, so that the newly expanded population contains the same genetic variation as the bottlenecked population (i.e. no new mutations during the bottleneck period). 

If my interpretation of your reply is correct, then perhaps I could achieve something similar to this by creating a parameter, t, for the length of the bottleneck, which is a fixed function of the length of the phylogeny. If t is fixed to be relatively short, then the number of new mutations occurring over the bottleneck period would be kept to a minimum. Would this work, and is there some minimum increment of time that I could use for t? Alternatively, is it possible to set the mutation probability to 0 during certain points of the simulation?

You can set the mutation rate to zero by setting theta0 = 0 for that part of the simulation.

There is no minimum increment of time in dadi. The effect of a bottleneck on diversity is proportional to t/N, so a very short sharp bottleneck (small t, small N) is equivalent to a longer less-sharp bottleneck (larger t, larger N), particularly if there's no new mutation.

Best,

Ryan

Regards

Ray

On Monday, February 4, 2013 12:37:10 PM UTC-8, Ryan Gutenkunst wrote:

Hello Ray,

I don't hink I entirely understand your question, so please correct me if this doesn't help.

If you want to model a bottleneck that recovers via a step-function, you would have three Integration.xxx_pop calls: one before the bottleneck, one during the bottleneck, and one after recovery. During each call you can use a different parameter for the population size.

A bottleneck with zero duration would have no effect on the SFS, because no drift would happen during that period.

Best,

Ryan

On 1/29/13 6:43 PM, "tinga...@gmail.com" <tinga...@gmail.com> wrote:

Hi

I'm a new dadi user interested in modeling a demography involving numerous population contractions and expansions. I would like to make the expansion phase following a contraction to be instantaneous rather than exponential. Based on what I have read in the manual and on this forum I am guessing that this could be achieved by setting the time of the bottleneck and the expansion to be the same, but allowing the population sizes in each event to be separately inferred. Would this work? Also how does one enforce explicit expansion or contraction in the Integration.xxx_pop function? 

Regards

Ray

-- 

Ryan Gutenkunst

Assistant Professor

Molecular and Cellular Biology

University of Arizona

phone: (520) 626-0569

http://gutengroup.mcb.arizona.edu

--
You received this message because you are subscribed to the Google Groups "dadi-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dadi-user+...@googlegroups.com.
To post to this group, send email to dadi...@googlegroups.com.
Visit this group at http://groups.google.com/group/dadi-user?hl=en-US.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[tinga...@gmail.com]

Hi Ryan

Thanks for the helpful feedback. I now have another question: I wish to admix two populations as a final step in my model. I have read in the manual that you can use from_phi to do so. I have tried to do this, but I keep getting the following error message:

ValueError: Dimensionality of phi and lengths of ns and xxs do not all agree.

I have used several different permutations of the arguments in the from_phi command in the final line of my code, including:

Spectrum.from_phi(phi, ns, (xx,xx), admix_props=((1-alpha,alpha),(0,1)))

Spectrum.from_phi(phi, ns, (xx,), admix_props=((1-alpha,alpha),(0,1)))

Spectrum.from_phi_2D_admix_props(nuNAm, nuNAm, xx, xx, phi, admix_props=((1-alpha,alpha),(0,1)))

etc, all to no avail.

Prior to the final step I have two populations, such that phi is two dimensional. Also, the code works fine if I admix the two populations prior to the last step (using phi_2D_to_3D_admix), then integrate out the two parental populations using trapz twice, and then get the sfs on the single admixed population using from_phi. However, doing this results in the warning message:

WARNING:Numerics:Extrapolation may have failed. Check resulting frequency spectrum for unexpected results.

Any ideas on where I am going wrong?

Regards

Ray

[tinga...@gmail.com]

Sorry, one more question. I have been getting an error message back from dadi.Inference.ll_multinom() stating:

    if data.folded and not model.folded:

AttributeError: 'function' object has no attribute 'folded'

or sometimes

AttributeError: 'MaskedArray' object has no attribute 'folded'

This is the first time that I have encountered this type of error message, and I am not sure what I have changed in my script to generate it. Nonetheless, the manual indicates that model spectra are automatically folded when used in conjunction with folded spectra. So it is strange that I am getting this error message at all.

Here are two spectra that generate the error (model is simulated fs):

>>> fs

Spectrum([-- 1371.37297581 1499.95429061 1605.38438404 1654.42894071 1648.46653169

 1598.01905034 1514.40576309 1423.57575161 1344.88329937 1275.68160363

 1213.44935962 1157.14711687 1109.45004351 1062.95827042 1022.79966234

 988.430682888 956.834293896 923.170887643 888.564185931 856.06611248

 822.949079176 790.187270171 755.862713291 722.991031193 689.772343275

 655.529911914 620.090585036 584.427179723 550.952116755 520.310698232

 493.25865533 474.249899867 464.724635697 -- -- -- -- -- -- -- -- -- -- --

 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --], folded=True, pop_ids=['FL_all'])

>>> model

Spectrum([-- 1.07546127927 0.494169769889 0.283225884585 0.175084415446

 0.111567500468 0.0714634383878 0.0451510713214 0.0277256052816

 0.0163844629711 0.00927549674982 0.00502752720019 0.00260823642757

 0.00128572251922 0.000585244725158 0.000225469755505 5.48963789957e-05

 1.55888567922e-06 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

 -- -- -- -- -- --], folded=False, pop_ids=None)

Note that if I try to fold the model spectrum manually, using model.fold, I get back the same spectrum, and if I try folding it with Spectrum.fold(model) I get an empty spectrum, so this doesn't help. Can you point me in the right direction?

[Gutenkunst, Ryan N - (rgutenk)]

Hello Ray,

That error comes from a check that the dimensionality of phi matches the length of ns and the length of the xx list. Is ns of length two in this example?

The admix_props argument assumes that you want an output fs that is the same dimensionality as phi, so in this case ns would need to be of length 2. It sounds like in the end you want only a single-poulation sfs, so you could just pass a placeholder for the sample size in the second population, ns = (ns[0], 1), the marginalize() the resulting spectrum.

The WARNING occurs when dadi detects that extrapolation has yielded a value in the sfs that is very different from the inputs used for extrapolation. This may be the result of some sort of numeric instability. Essentially, it's warning you that you should double-check the output FS and ensure it is sensible.

Best,

Ryan

[Gutenkunst, Ryan N - (rgutenk)]

Hello Ray,

It's unclear from your email what's happening. Can you send a code sample that generates the errors you're seeing?

The first error appears to result from passing a function in place of model or data Spectrum object into ll_multinom. It's not clear to me how that might be happening, as it might be an error in your upstream code.

The second error might happen if a Spectrum is somehow downgraded to a basic array. I'm not sure where that might be happening in your code.

Best,

Ryan

[Ray Tobler]

Hi Ryan

Thanks for the prompt reply. I have attached a version of the code that generates the error message (specifically: AttributeError: 'MaskedArray' object has no attribute 'folded') along with a sample of my dataset.

Many thanks for taking the time to look into this for me. Your ongoing help is very much appreciated.

Cheers

Ray

[Gutenkunst, Ryan N - (rgutenk)]

Hi Ray,

The problem was that you have "phi = PhiManip.phi_1D_to_2D(phi, xx)" when you should have "phi = PhiManip.phi_1D_to_2D(xx, phi)". I've added a check to the dadi SVN code to address it.

To track this down, I looked at the model fs you were getting after extrapolation, and I saw that it was already masked at high frequencies. Calculating an SFS without extrapolation, I saw that these entries were negative (which causes nans when extrapolating using logs.) Tracking deeper, I saw that in your demographic function the intermediate phis had large negative values. This eventually led me to early in your demographic function, where I saw the bug.

While tracking this down, I also discovered a bug that prevents use of admix_props in from_phi for 2D models. That has also been fixed in the SVN code base.

Best,

Ryan

[Ray Tobler]

Hi Ryan

Funnily enough I was just checking this when I got your reply. I must have unwittingly switched the argument order around when I was modifying my script. Everything seems to be working now, so thanks once again for your excellent help!

Cheers

Ray

[Ray Tobler]

Hi Ryan

OK, everything has been working smoothly thanks to your suggestions. However, I have now modified one of my models to include three populations, with the intention of using the Spectrum.from_phi() function to admix two of the populations at the end. Now, if I do not include the admix_props argument (i.e. leave it at the default settings) then the model runs without any problems. However, if I set admix_props to ((1,0,0),(0,1,0),(alpha,0,1-alpha)), which I assume would admix the 3rd population with a proportion alpha from the 1st population, then I get the following error message:

Traceback (most recent call last):

  File "<stdin>", line 1, in <module>

  File "<stdin>", line 5, in runModel

  File "/usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg/dadi/Numerics.py", line 319, in extrap_func

    result_l = map(partial_func, pts_l)

  File "<stdin>", line 16, in admix_end_func

  File "/usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg/dadi/Spectrum_mod.py", line 1359, in from_phi

    admix_props, het_ascertained)

  File "/usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg/dadi/Spectrum_mod.py", line 1221, in _from_phi_3D_direct

    ans = numpy.sum(half_dx * (integrand[1:]+integrand[:-1]))

ValueError: operands could not be broadcast together with shapes (39) (39,40,40) 

In seeing this I then attempted to use the function  _from_phi_3D_admix_props in place of from_phi, but I got an error message stating that Spectrum had no such attribute (notably it does not appear on the dir() list for Spectrum either).

I have attached the new code that generates the error message. Sorry to be such a bother, but as you can probably gather I am still a novice when it comes to these things.

Cheers

Ray

[Gutenkunst, Ryan N - (rgutenk)]

Hi Ray,

I suspect this is due to a bug that is fixed in the current SVN version. (In part, because I don't get that error when I run your code.) Are you comfortable following these instructions ( https://code.google.com/p/dadi/source/checkout ) to grab the current source distribution and then installing it? If not, what system are you running on, and I can send you a binary.

Best,

Ryan

[Ray Tobler]

Hi Ryan 

I have checked out the new version of dadi (v 480), following the instructions on the website (i.e. svn checkout http://dadi.googlecode.com/svn/trunk/ dadi-read-only). However, now when I attempt to import dadi, I get the following error message:

Traceback (most recent call last):

  File "<stdin>", line 1, in <module>

  File "dadi/__init__.py", line 11, in <module>

    import Demographics1D

  File "dadi/Demographics1D.py", line 6, in <module>

    from dadi import Numerics, PhiManip, Integration

  File "dadi/Integration.py", line 21, in <module>

    import Misc, Numerics, tridiag

ImportError: No module named tridiag

Have I not done something I should have?

Cheers

Ray

[Gutenkunst, Ryan N - (rgutenk)]

You'll need to install the source code. Following the instructions in section 9.3 of the manual.

Best,

Ryan

On 2/21/13 11:16 AM, "Ray Tobler" <tinga...@gmail.com> wrote

[Ray Tobler]

Hi Ryan

I am clearly not doing this correctly, as I am still getting the same error message.

First, I updated dadi using svn checkout http://dadi.googlecode.com/svn/trunk/ dadi-read-only

Next, following the instructions in the manual, I ran sudo setup python setup.py install

I then reran the code that I attached in the previous email, and got the same error message as before:

Traceback (most recent call last):

  File "<stdin>", line 1, in <module>

  File "<stdin>", line 16, in admix_end_func

  File "/usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg/dadi/Spectrum_mod.py", line 1359, in from_phi

    admix_props, het_ascertained)

  File "/usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg/dadi/Spectrum_mod.py", line 1221, in _from_phi_3D_direct

    ans = numpy.sum(half_dx * (integrand[1:]+integrand[:-1]))

ValueError: operands could not be broadcast together with shapes (39) (39,40,40)

Do I need to reinstall dadi from scratch?

Cheers

Ray

[Ray Tobler]

correction: 

That should read

 sudo python setup.py install

and not

sudo setup python setup.py install

[Gutenkunst, Ryan N - (rgutenk)]

Hi Ray,

You're still hitting the old version. What messages are you getting at the end of the python setup.py install command? You should see lines indicating where the files are being installed.

Best,

Ryan

[Ray Tobler]

Hi Ryan

I guess there could be more than one version of dadi floating around on the server I am using to run the simulations. However, I couldn't find any evidence when searching for it using find, which etc.

Below is the full output after running the setup.py script:

running install

running bdist_egg

running egg_info

running build_src

build_src

building extension "dadi.tridiag" sources

f2py options: []

  adding 'build/src.linux-x86_64-2.7/fortranobject.c' to sources.

  adding 'build/src.linux-x86_64-2.7' to include_dirs.

building extension "dadi.integration_c" sources

f2py options: []

  adding 'build/src.linux-x86_64-2.7/fortranobject.c' to sources.

  adding 'build/src.linux-x86_64-2.7' to include_dirs.

build_src: building npy-pkg config files

writing dadi.egg-info/PKG-INFO

writing top-level names to dadi.egg-info/top_level.txt

writing dependency_links to dadi.egg-info/dependency_links.txt

reading manifest file 'dadi.egg-info/SOURCES.txt'

reading manifest template 'MANIFEST.in'

warning: no previously-included files matching '*.msout' found anywhere in distribution

warning: no previously-included files matching '*.swp' found anywhere in distribution

warning: no previously-included files matching 'seedms' found anywhere in distribution

warning: no previously-included files matching '*.png' found anywhere in distribution

writing manifest file 'dadi.egg-info/SOURCES.txt'

installing library code to build/bdist.linux-x86_64/egg

running install_lib

running build_py

copying dadi/dadi_models_hardcode_fixedparams_3sfs.py -> build/lib.linux-x86_64-2.7/dadi

running build_ext

customize UnixCCompiler

customize UnixCCompiler using build_ext

running scons

creating build/bdist.linux-x86_64/egg

creating build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Numerics.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Spectrum_mod.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Misc.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Integration.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/integration_c.so -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/RunInParallel.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/svnversion -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/tridiag.so -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Inference.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Plotting.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Demographics2D.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Demographics1D.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/__init__.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/Hessian.py -> build/bdist.linux-x86_64/egg/dadi

copying build/lib.linux-x86_64-2.7/dadi/PhiManip.py -> build/bdist.linux-x86_64/egg/dadi

byte-compiling build/bdist.linux-x86_64/egg/dadi/Numerics.py to Numerics.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Spectrum_mod.py to Spectrum_mod.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Misc.py to Misc.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Integration.py to Integration.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/RunInParallel.py to RunInParallel.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Inference.py to Inference.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Plotting.py to Plotting.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Demographics2D.py to Demographics2D.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Demographics1D.py to Demographics1D.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/dadi_models_hardcode_fixedparams_3sfs.py to dadi_models_hardcode_fixedparams_3sfs.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/__init__.py to __init__.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/Hessian.py to Hessian.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/PhiManip.py to PhiManip.pyc

creating stub loader for dadi/tridiag.so

creating stub loader for dadi/integration_c.so

byte-compiling build/bdist.linux-x86_64/egg/dadi/tridiag.py to tridiag.pyc

byte-compiling build/bdist.linux-x86_64/egg/dadi/integration_c.py to integration_c.pyc

creating build/bdist.linux-x86_64/egg/EGG-INFO

installing scripts to build/bdist.linux-x86_64/egg/EGG-INFO/scripts

running install_scripts

running build_scripts

creating build/bdist.linux-x86_64/egg/EGG-INFO/scripts

copying build/scripts.linux-x86_64-2.7/ms_jsfs.py -> build/bdist.linux-x86_64/egg/EGG-INFO/scripts

changing mode of build/bdist.linux-x86_64/egg/EGG-INFO/scripts/ms_jsfs.py to 755

copying dadi.egg-info/PKG-INFO -> build/bdist.linux-x86_64/egg/EGG-INFO

copying dadi.egg-info/SOURCES.txt -> build/bdist.linux-x86_64/egg/EGG-INFO

copying dadi.egg-info/dependency_links.txt -> build/bdist.linux-x86_64/egg/EGG-INFO

copying dadi.egg-info/top_level.txt -> build/bdist.linux-x86_64/egg/EGG-INFO

writing build/bdist.linux-x86_64/egg/EGG-INFO/native_libs.txt

zip_safe flag not set; analyzing archive contents...

dadi.__init__: module references __file__

creating 'dist/dadi-1.6.3-py2.7-linux-x86_64.egg' and adding 'build/bdist.linux-x86_64/egg' to it

removing 'build/bdist.linux-x86_64/egg' (and everything under it)

Processing dadi-1.6.3-py2.7-linux-x86_64.egg

removing '/usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg' (and everything under it)

creating /usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg

Extracting dadi-1.6.3-py2.7-linux-x86_64.egg to /usr/lib64/python2.7/site-packages

dadi 1.6.3 is already the active version in easy-install.pth

Installing ms_jsfs.py script to /usr/bin

Installed /usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg

Processing dependencies for dadi==1.6.3

Finished processing dependencies for dadi==1.6.3

Not sure what I'm looking for, but both the installation and error message appear to be using /usr/lib64/python2.7/site-packages/dadi-1.6.3-py2.7-linux-x86_64.egg.

Cheers

Ray

[Gutenkunst, Ryan N - (rgutenk)]

Hi Ryan,

I'm not entirely sure what's going on. Try manually deleting everything dadi-related in /usr/lib64/python2.7/site-packages and the running the installer again. Perhaps it's not overwriting what it thinks is a current installation. You can check for success if you end up with the version of Spectrum_mod that is here: https://code.google.com/p/dadi/source/browse/trunk/dadi/Spectrum_mod.py . Also ensure that the copy in the install directory matches this online version. (You can just look at line 1221, which was the offending line in your previous error.)

Best,

Ryan

[Ray Tobler]

Hi Ryan

I have deleted everything in /usr/lib64/python2.7/site-packages (i.e. the egg directory), and rerun the installer, but got the same error message. I have tried re-installing dadi from scratch, and doing a clean install on my own laptop, but still no dice. Notably, when I check the Spectrum module, neither the 2D nor 3D from_phi_XX_admix_props function is present, so it looks like I am somehow ending up with the older version. Might there be some issue with the updating at your end?

Cheers

Ray

[Gutenkunst, Ryan N - (rgutenk)]

Hi Ray,

Everything looks okay on my end. Have you tried re-checking-out the code from SVN? Did you check that the code you're installing matches what is currently in the SVN repository, by following the link I gave you? If you're still having trouble with checkout, try the attached source .zip.

Best,

Ryan