I am trying to create a network using a glif_cond model in PointNet, as I already did using a glif_asc model previously (and which worked well).
To do so, I followed the same procedure described in your tutorial, which is by using json configuration file downloaded from the Allen Cell Type Database. However, when I try to create the network using glif_cond models (specifying glif_cond in the add_nodescall, I get the following error :
sim = pointnet.PointSimulator.from_config(configure, network)
2024-06-27 15:06:52,685 [INFO] Setting up output directory
2024-06-27 15:06:52,686 [INFO] Building cells.
/Users/julienballbe/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/nest/lib/hl_api_nodes.py:151: UserWarning:
SetStatus() call failed, but nodes have already been created! The node IDs of the new nodes are: NodeCollection(metadata=None, model=glif_cond, size=19, first=1, last=19).
Traceback (most recent call last):
Cell In[65], line 1
sim = pointnet.PointSimulator.from_config(configure, network)
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/bmtk/simulator/pointnet/pointsimulator.py:278 in from_config
graph.build_nodes()
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/bmtk/simulator/pointnet/pointnetwork.py:152 in build_nodes
node.build()
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/bmtk/simulator/pointnet/sonata_adaptors.py:65 in build
self._nest_objs = nest.Create(self.nest_model, self.n_nodes, self.nest_params)
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/nest/ll_api.py:216 in stack_checker_func
return f(*args, **kwargs)
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/nest/lib/hl_api_nodes.py:149 in Create
SetStatus(node_ids, params)
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/nest/ll_api.py:216 in stack_checker_func
return f(*args, **kwargs)
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/nest/lib/hl_api_info.py:315 in SetStatus
sli_func("SetStatus", nodes, params)
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/nest/ll_api.py:152 in sli_func
sli_run(slifun) # SLI support code to execute func on args
File ~/anaconda3/envs/wBMTKAllenSDKNEST/lib/python3.11/site-packages/nest/ll_api.py:103 in catching_sli_run
raise exceptionCls(commandname, message)
NESTErrors.DictError: DictError in SLI function SetStatus_id: Unused dictionary items: dt type El_reference C asc_amp_array init_threshold threshold_reset_method th_inf spike_cut_length init_AScurrents init_voltage threshold_dynamics_method voltage_reset_method extrapolation_method_name voltage_dynamics_method El asc_tau_array R_input AScurrent_dynamics_method AScurrent_reset_method dt_multiplier th_adapt coeffs
Even though the same script works if I specify "glif_psc" model in the add_nodescall.
I tried to create two neurons directly in NEST (one current based using glif_psc, and one conductance based using glif_cond), and I get the same error for both.
Do you know where the error comes from? I guess it is because the parameters specified in the json configuration files of models (El_ref, C, ...) do not appear under the same denomination in the glif_psc/glif_conf cpp files
Is there in the BMTK some shortcuts which convert parameters names in the json configuration file for glif_psc models?
Hi,
I am trying to create a network using a glif_cond model in PointNet, as I already did using a glif_asc model previously (and which worked well). To do so, I followed the same procedure described in your tutorial, which is by using json configuration file downloaded from the Allen Cell Type Database. However, when I try to create the network using glif_cond models (specifying glif_cond in the
add_nodescall, I get the following error :Even though the same script works if I specify "glif_psc" model in the
add_nodescall.I tried to create two neurons directly in NEST (one current based using glif_psc, and one conductance based using glif_cond), and I get the same error for both. Do you know where the error comes from? I guess it is because the parameters specified in the json configuration files of models (El_ref, C, ...) do not appear under the same denomination in the glif_psc/glif_conf cpp files Is there in the BMTK some shortcuts which convert parameters names in the json configuration file for glif_psc models?
Thank you for any help you can give me,
Best,
Julien