Epicodim3 #70

LenSpek · 2017-11-09T16:58:41Z

My implementation of the Epileptor Codim 3 model by Marisa.
(Ignore the first 3 commits, they were added by accident)

A python implementation of Marisa's codim 3 Epileptor model for TVB

Updating local trunk

Fixed a bug in the ultra-slow modulation and improved some documentation

Adjusted the packages to be in line with the other TVB-models

maedoc · 2017-11-09T23:44:32Z

tvb/simulator/models/epileptorcodim3.py

+        High Codimension Singularity and the Ultra-slow Transitions of Classes.*
+        Journal of Mathematical Neuroscience. 2017;7:7. doi:10.1186/s13408-017-0050-8.
+
+    .. The Epileptor codim 3 model is a neural mass model which contains two subsystems acting at different timescales.


please reflow to 80 columns, in Vim this is gqip but your editor may be different. or I can do it for you.

I will adjust this

maedoc · 2017-11-09T23:46:07Z

tvb/simulator/models/epileptorcodim3.py

+        self.F = self.F / numpy.linalg.norm(self.F)
+
+
+class EpileptorCodim3_slowmod(Model):


Is this a completely different model?

This model is similar, it is a different implementation. It allows for multiple classes of bursters to be visited by a simulation. The model dynamics are essentially the same. However there are some differences in the input of the system and some extra dynamics to allow the transition of classes

maedoc · 2017-11-09T23:49:35Z

@LenSpek looking good; I have a few requests:

keep the dfuns as you have them but try to add numba versions, as in the epileptor 6d model, and use it as the dfun. please test & eventually commit to this branch/PR
please provide some reproductions of figures from the paper either as a Python script or prefereably as a Jupyter notebook

Reflowed the documentation to a column with of 80 and improved the documentation of the _slowmod implementation

maedoc

Looking good! 👍🏼
Careful though, if you implement Numba dfun for second model, you should change the names to be unique .

maedoc · 2017-11-15T20:56:19Z

How much time do you have left to work in this model, btw?

LenSpek · 2018-01-10T15:28:36Z

I have time till the end of January. However if there are a few minor things, I will be able to address these after my internship has ended

maedoc · 2018-01-12T13:33:33Z

thanks and sorry this took so long.

liadomide · 2018-01-12T15:03:41Z

After this recent merge, one unit-test fails:

tvb-library/tvb/tests/simulator/simulator_test.py: TestSimulator.test_simulator_region

Error Message
TypeError: ufunc '_numba_dfun' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''
....
../scientific_library/tvb/simulator/models/epileptorcodim3.py:286: TypeError

Can you please check if the test was wrongly written, or the new model is missing some API implicit requirement ?

maedoc · 2018-01-12T15:06:03Z

The test probably assumes all parameters are floats while some parameters in this model are integers, and it's unsafe to convert a float to an integer.

maedoc · 2018-01-12T15:08:08Z

tvb/simulator/models/epileptorcodim3.py

+        """"The dfun using numba for speed"""
+        state_variables_ = state_variables.reshape(state_variables.shape[:-1]).T
+        coupling_ = coupling.reshape(coupling.shape[:-1]).T
+        derivative = _numba_dfun(state_variables_, coupling_, self.E[0], self.E[1], self.E[2], self.F[0], self.F[1],


@LenSpek can you set a breakpoint here and double check that the types of those non-float parameters being passed to the numba dfun match or if not, cast their values correclty?

I double-checked and the variables in question (self.modification, self.N) have dtype int32.

is that the case for defaults as well (when you don't specify any model parameters)? this test runs your model as if you made a simulation without specifying any parameters, only defaults

LenSpek · 2018-01-18T16:07:08Z

I think that I found what was triggering the tests: Modification is specified by default as a BoolArray, which TVB converts to an int32. I will push a change to address this, so the test works.
(BTW, when I run the TVB tests, I don't get this error)

liadomide · 2018-01-18T17:28:26Z

If i debug in my env, the 2 params which are expected to be int32 are in fact int64.
I get the same error both in unit-test and in the ipython notebook that came with these new models.
I suspect this is the default of the installation.

LenSpek · 2018-01-18T17:41:22Z

I find that strange as I use the TVB classes from array.py and there the class IntegerArray is defined as:

class IntegerArray(BaseArray):
    _ui_name = "Array of integers"
    dtype = basic.DType(default=numpy.int32)

liadomide · 2018-01-19T08:33:50Z

Could we use dtype int_ ?
This would become int32 on Windows OS and int64 on Unix systems.

I tested on Anaconda x64 with Python 2.7.14 and numpy 1.13 on Linux, Mac and Windows. If we replace the current int32 with int_, then all 3 systems are ok. Otherwise, with the current code only Windows is fine, the others are not working.

maedoc · 2018-01-19T08:53:25Z

@liadomide can you paste a git diff of your int_ change, so it's clear where the change is made?

maedoc · 2018-01-19T08:55:05Z

Numba supports the dtypes in question, and I would prefer platform independence if possible. (int_ corresponds to C long)

liadomide · 2018-01-19T09:05:26Z

diff.txt

liadomide · 2018-01-19T10:41:40Z

int_ is the default Integer type for Numpy, that is why I suggest it.
Seems to work well with numba int_
https://docs.scipy.org/doc/numpy-1.13.0/user/basics.types.html

maedoc · 2018-01-19T10:56:06Z

OK then, int_ it is.

liadomide · 2018-01-19T10:59:21Z

Good. I will then apply this change soon.

LenSpek added 8 commits October 17, 2017 15:56

Epileptor codim 3 model

ccc1f4d

A python implementation of Marisa's codim 3 Epileptor model for TVB

Made a commit in the wrong branch

6c7fa5d

Merge pull request #1 from the-virtual-brain/trunk

f6290df

Updating local trunk

First version of epileptor codim 3 model

64f7a3a

Added the slow-modulation version of the Epileptor Codim 3 model

93e32ec

Bugfixing and documentation

5812233

Fixed a bug in the ultra-slow modulation and improved some documentation

Included EpileptorCodim3 in init.py

f326705

Simplified the packages

05fcc7b

Adjusted the packages to be in line with the other TVB-models

maedoc reviewed Nov 9, 2017

View reviewed changes

LenSpek added 2 commits November 10, 2017 16:34

Improved documentation

7685171

Reflowed the documentation to a column with of 80 and improved the documentation of the _slowmod implementation

Added the numba dfun version for the Epileptor Codim 3 model

d941db7

maedoc reviewed Nov 15, 2017

View reviewed changes

Added the numba version for the ultra slow modulation

22f47dc

maedoc merged commit 0eddbf9 into the-virtual-brain:trunk Jan 12, 2018

maedoc reviewed Jan 12, 2018

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Epicodim3 #70

Epicodim3 #70

LenSpek commented Nov 9, 2017

maedoc Nov 9, 2017

LenSpek Nov 10, 2017

maedoc Nov 9, 2017

LenSpek Nov 10, 2017

maedoc commented Nov 9, 2017

maedoc left a comment

maedoc commented Nov 15, 2017

LenSpek commented Jan 10, 2018

maedoc commented Jan 12, 2018

liadomide commented Jan 12, 2018

maedoc commented Jan 12, 2018

maedoc Jan 12, 2018

LenSpek Jan 12, 2018

maedoc Jan 12, 2018

LenSpek commented Jan 18, 2018

liadomide commented Jan 18, 2018 •

edited

LenSpek commented Jan 18, 2018 •

edited by maedoc

liadomide commented Jan 19, 2018

maedoc commented Jan 19, 2018

maedoc commented Jan 19, 2018 •

edited

liadomide commented Jan 19, 2018

liadomide commented Jan 19, 2018

maedoc commented Jan 19, 2018

liadomide commented Jan 19, 2018

		self.F = self.F / numpy.linalg.norm(self.F)


		class EpileptorCodim3_slowmod(Model):

Epicodim3 #70

Epicodim3 #70

Conversation

LenSpek commented Nov 9, 2017

maedoc Nov 9, 2017

Choose a reason for hiding this comment

LenSpek Nov 10, 2017

Choose a reason for hiding this comment

maedoc Nov 9, 2017

Choose a reason for hiding this comment

LenSpek Nov 10, 2017

Choose a reason for hiding this comment

maedoc commented Nov 9, 2017

maedoc left a comment

Choose a reason for hiding this comment

maedoc commented Nov 15, 2017

LenSpek commented Jan 10, 2018

maedoc commented Jan 12, 2018

liadomide commented Jan 12, 2018

maedoc commented Jan 12, 2018

maedoc Jan 12, 2018

Choose a reason for hiding this comment

LenSpek Jan 12, 2018

Choose a reason for hiding this comment

maedoc Jan 12, 2018

Choose a reason for hiding this comment

LenSpek commented Jan 18, 2018

liadomide commented Jan 18, 2018 • edited

LenSpek commented Jan 18, 2018 • edited by maedoc

liadomide commented Jan 19, 2018

maedoc commented Jan 19, 2018

maedoc commented Jan 19, 2018 • edited

liadomide commented Jan 19, 2018

liadomide commented Jan 19, 2018

maedoc commented Jan 19, 2018

liadomide commented Jan 19, 2018

liadomide commented Jan 18, 2018 •

edited

LenSpek commented Jan 18, 2018 •

edited by maedoc

maedoc commented Jan 19, 2018 •

edited