A question about LIF model in pynn

[Qianhui Liu]

Hi:

   I am translating code wrote by Brain1 to Pynn/Nest. Now I have a question about it and seek help.

   The question is, I define a neuron which is LIF model. Then define a input neuron which spikes at 10ms and connects to the previous one. I define the same condition in Brian1 and Pynn/Nest (also the same weight).  But it turns out that the neuron in Brian fires while the neuron in Pynn does not. I almost make sure that the parameter  I give in Pynn is correct, but still got the different result. 

Pynn/Nest result 

Brian result

[Shailesh Appukuttan]

Hi,

The weights assigned to the inputs in the two cases are not equal:

In Brian:

connections['XeAe'].connect(input_groups['e'], neuron_groups['e'], 1)

In PyNN:

connections['XeAe'] = sim.Projection(input_groups['e'], neuron_groups['e'], sim.OneToOneConnector(),sim.StaticSynapse(weight=1))

Brian considers the value of "1" as "1 s" while PyNN takes it as "1 ms" as is their normal convention.

Setting a strength of "1 s" might be too high a value. So we suggest you set it to a lower value.

Also, do verify that your neuron equations are correct.

Cheers.

[Qianhui Liu]

Thank u very much! I am too careless.

Now I try to verify the parameters I given in Pynn. But I still puzzle by the parameter 'cm' and 'tan_m' because of the difference between the unit in Pynn and Nest. The 'cm' in pynn is nF, the corresponding parameter in nest is 'C_m' which unit is pF.The 'tan_m' in pynn is nF, the corresponding parameter in nest is 'g_l' which unit is pF. So I am not sure that the parameter I given in pynn is equals to the equations in Brian. Could you please have a look at this?
it
在 2017年4月1日星期六 UTC+8上午12:44:50，Shailesh Appukuttan写道：

[Shailesh Appukuttan]

Hi, I am not really sure what you meant by the parameter "tan_m"? I couldn't find it anywhere in your code.

tan_m is not a parameter within PyNN. Are you referring to "tau_m"?.. if so, the units would be of time (ms).