from matplotlib import pyplot as plt
import sys
nml2_file = 'NML2_SingleCompHHCell.nml'
target = "net1"
dts = [0.1, 0.05, 0.01, 0.005, 0.001, 0.0005, 0.0001]
dts = [0.05, 0.01, 0.005]
nogui = '-nogui' in sys.argv # Used to supress GUI in tests for Travis-CI
analyse_spiketime_vs_dt(nml2_file,
target,
300,
'jNeuroML',
'hhpop[0]/v',
dts,
verbose=True,
show_plot_already=False)
if not nogui:
analyse_spiketime_vs_dt(nml2_file,
target,
300,
'jNeuroML_NEURON',
'hhpop[0]/v',
dts,
verbose=True,
show_plot_already=False)
if not nogui:
plt.show()
def dashboard_cells(net_id,
net_file_name,
config_array,
global_dt,
if_params,
elec_len_list,
dt_list,
generate_dashboards=True,
compare_to_neuroConstruct=False,
regenerate_nml2=False,
show_plot_already=False,
proj_string_neuroConstruct=None,
shell=None,
nc_home=None):
################### check whether use of neuroConstruct python/java interface is needed ########################
if regenerate_nml2:
use_NeuroConstruct(compare_to_neuroConstruct=compare_to_neuroConstruct,
regenerate_nml2=regenerate_nml2,
proj_string=proj_string_neuroConstruct,
global_dt=global_dt,
config_array=config_array,
elec_len_list=elec_len_list,
shell=shell,
nc_home=nc_home)
else:
use_NeuroConstruct(compare_to_neuroConstruct=compare_to_neuroConstruct,
regenerate_nml2=regenerate_nml2,
proj_string=proj_string_neuroConstruct,
global_dt=global_dt,
config_array=config_array,
shell=shell,
nc_home=nc_home)
############################################################################################################
if generate_dashboards:
cell_id_list=[]
config_list=[]
analysis_header_list=[]
nC_vs_NML2Curve_list=[]
IFcurve_list=[]
for cellModel in config_array.keys():
cell_id_list.append(cellModel)
config_list.append(config_array[cellModel]['Analysis'])
save_to_path="../"+cellModel
if not os.path.exists(save_to_path):
print("Creating a new directory %s"%save_to_path)
os.makedirs(save_to_path)
else:
print("A directory %s already exists"%save_to_path)
pathToConfig="../"+config_array[cellModel]['Analysis']
try:
with open(os.path.join(pathToConfig,"compSummary.json"),'r') as f:
comp_info=json.load(f)
except IOError:
print "cannot open file %s"%os.path.join(pathToConfig,"compSummary.json")
cell_morph_summary=comp_info[config_array[cellModel]['Analysis']]
src_files = os.listdir(pathToConfig)
num_dx_configs=0
dx_array=[]
found_all_compartmentalizations=False
dx_configs={}
target_v=None
found_default=False
for file_name in src_files:
full_file_path=os.path.join(pathToConfig,file_name)
if (os.path.isdir(full_file_path)) and "_default" in file_name:
found_default=True
original_LEMS_target=os.path.join(full_file_path,"LEMS_Target.xml")
###################################################################################
if -1 in elec_len_list and "-1" in cell_morph_summary.keys():
dx_configs[cell_morph_summary["-1"]["IntDivs"]]=original_LEMS_target
default_num_of_comps=cell_morph_summary["-1"]["IntDivs"]
dx_array.append(int(default_num_of_comps))
num_dx_configs+=1
##################################################################################
print("%s is a directory"%full_file_path)
print("will generate the IF curve for %s.cell.nml"%cellModel)
generate_current_vs_frequency_curve(os.path.join(full_file_path,cellModel+".cell.nml"),
cellModel,
start_amp_nA = if_params['start_amp_nA'],
end_amp_nA = if_params['end_amp_nA'],
step_nA = if_params['step_nA'],
analysis_duration =if_params['analysis_duration'],
analysis_delay = if_params['analysis_delay'],
dt= if_params['dt'],
temperature= if_params['temperature'],
plot_voltage_traces=if_params['plot_voltage_traces'],
plot_if= if_params['plot_if'],
plot_iv= if_params['plot_iv'],
show_plot_already= show_plot_already,
save_if_figure_to='%s/IF_%s.png'%(save_to_path,cellModel),
save_iv_figure_to='%s/IV_%s.png'%(save_to_path,cellModel),
simulator= if_params['simulator'])
IFcurve="IF_%s"%cellModel
IFcurve_list.append(IFcurve)
IVcurve="IV_%s"%cellModel
nml2_file_path=os.path.join(full_file_path,net_file_name+".net.nml")
net_doc = pynml.read_neuroml2_file(nml2_file_path)
net=net_doc.networks[0]
pop=net.populations[0]
popID=pop.id
target_v="%s/0/%s/v"%(popID,cellModel)
########################################################################################
if if_params['simulator'] == 'jNeuroML':
results = pynml.run_lems_with_jneuroml(original_LEMS_target, nogui=True, load_saved_data=True, plot=False, verbose=False)
if if_params['simulator'] == 'jNeuroML_NEURON':
results = pynml.run_lems_with_jneuroml_neuron(original_LEMS_target, nogui=True, load_saved_data=True, plot=False, verbose=False)
t = results['t']
v = results[target_v]
if compare_to_neuroConstruct:
print("will generate the comparison between the nC model and NeuroML2 model")
PlotNC_vs_NML2({'NML2':[{'t':t,'v':v}],'nC':[config_array[cellModel]['OriginalTag']],
'subplotTitles':['NeuroML2 versus nC model: simulations in NEURON with dt=%f'%global_dt]},
{'cols':8,'rows':5},
legend=True,
show=False,
save_to_file='%s/nC_vs_NML2_%s.png'%(save_to_path,config_array[cellModel]['Analysis']),
nCcellPath=os.path.join(save_to_path,config_array[cellModel]['Analysis']) )
analysis_string1="nC_vs_NML2_%s"%config_array[cellModel]['Analysis']
analysis_header1="Comparison between the original nC model and NeuroML2 model: simulations in NEURON with dt=%f"%global_dt
analysis_header_list.append(analysis_header1)
nC_vs_NML2Curve_list.append(analysis_string1)
else:
print("will generate the plot for the NeuroML2 model")
generate_nml2_plot({'NML2':[{'t':t,'v':v}],'subplotTitles':['NeuroML2 model: simulations in NEURON with dt=%f'%global_dt]},
{'cols':8,'rows':5},
show=False,
save_to_file='%s/NML2_%s.png'%(save_to_path,config_array[cellModel]['Analysis']))
analysis_string1="NML2_%s"%config_array[cellModel]['Analysis']
analysis_header1="NeuroML2 model: simulations in NEURON with dt=%f"%global_dt
analysis_header_array.append(analysis_header1)
nC_vs_NML2Curve_array.append(analysis_string1)
smallest_dt=min(dt_list)
########################################################################################
print("will generate the spike times vs dt curve for %s.cell.nml"%cellModel)
analyse_spiketime_vs_dt(nml2_file_path,
net_id,
get_sim_duration(os.path.join(full_file_path,"LEMS_%s.xml"%net_id)),
if_params['simulator'],
target_v,
dt_list,
verbose=False,
spike_threshold_mV = 0,
show_plot_already=show_plot_already,
save_figure_to="%s/Dt_%s.png"%(save_to_path,cellModel),
num_of_last_spikes=None)
dt_curve="Dt_%s"%cellModel
if not found_all_compartmentalizations:
for elecLen in range(0,len(elec_len_list)):
elec_len=str(elec_len_list[elecLen])
if elec_len != "-1":
if (elec_len in file_name) and (elec_len in cell_morph_summary.keys() ):
dx_configs[cell_morph_summary[elec_len]["IntDivs"]]=os.path.join(full_file_path,"LEMS_Target.xml")
num_dx_configs+=1
dx_array.append(int(cell_morph_summary[elec_len]["IntDivs"] ) )
break
if num_dx_configs==len(elec_len_list):
found_all_compartmentalizations=True
if not found_default:
print("default configuration for %s analysis is not found; execution will terminate; set regenerate_nml2 to True to generate target dirs."%cellModel)
quit()
if found_all_compartmentalizations:
dx_array=list(np.sort(dx_array) )
maximum_int_divs=max(dx_array)
print("testing the presence of cell configs with different levels of spatial discretization")
analyse_spiketime_vs_dx(dx_configs,
if_params['simulator'],
target_v,
verbose=False,
spike_threshold_mV = 0,
show_plot_already=show_plot_already,
save_figure_to="%s/Dx_%s.png"%(save_to_path,cellModel),
num_of_last_spikes=3)
dx_curve="Dx_%s"%cellModel
else:
print("not all of the target configurations were recompartmentalized; execution will terminate; set regenerate_nml2 to True to obtain all of the target configurations.")
quit()
if config_array[cellModel]['Analysis'] != config_array[cellModel]['SpikeProfile']:
pathToProfileConfig="../"+config_array[cellModel]['SpikeProfile']+"/"+config_array[cellModel]['SpikeProfile']+"_default"
original_LEMS_target=os.path.join(pathToProfileConfig,"LEMS_Target.xml")
if if_params['simulator'] == 'jNeuroML':
results = pynml.run_lems_with_jneuroml(original_LEMS_target, nogui=True, load_saved_data=True, plot=False, verbose=False)
if if_params['simulator'] == 'jNeuroML_NEURON':
results = pynml.run_lems_with_jneuroml_neuron(original_LEMS_target, nogui=True, load_saved_data=True, plot=False, verbose=False)
if 'SpikeProfileTag' in config_array[cellModel].keys():
tag=config_array[cellModel]['SpikeProfileTag']+"_0_wtime"
else:
tag=config_array[cellModel]['OriginalTag']
if 'SpikeProfileCellTag' in config_array[cellModel].keys() and 'SpikeProfileTag' in config_array[cellModel].keys():
target_v="%s/0/%s/v"%(config_array[cellModel]['SpikeProfileTag'],config_array[cellModel]['SpikeProfileCellTag'])
t = results['t']
v = results[target_v]
if compare_to_neuroConstruct:
print("will generate the comparison between the nC model and NeuroML2 model")
PlotNC_vs_NML2({'NML2':[{'t':t,'v':v}],'nC':[tag],
'subplotTitles':['NML2 versus nC model: simulations in NEURON with dt=%f'%global_dt]},
{'cols':8,'rows':5},
legend=True,
show=show_plot_already,
save_to_file='%s/nC_vs_NML2_%s.png'%(save_to_path,config_array[cellModel]['SpikeProfile']),
nCcellPath=os.path.join(save_to_path,config_array[cellModel]['SpikeProfile']) )
analysis_string2="nC_vs_NML2_%s"%config_array[cellModel]['SpikeProfile']
analysis_header2="Comparison between the original nC model and NeuroML2 model: simulations in NEURON with dt=%f"%global_dt
else:
print("will generate the plot for the NeuroML2 model")
generate_nml2_plot({'NML2':[{'t':t,'v':v}],'subplotTitles':['NeuroML2 model: simulations in NEURON with dt=%f'%global_dt]},
{'cols':8,'rows':5},
show=False,
save_to_file='%s/NML2_%s.png'%(save_to_path,config_array[cellModel]['SpikeProfile']))
analysis_string2="NML2_%s"%config_array[cellModel]['SpikeProfile']
analysis_header2="NeuroML2 model: simulations in NEURON with dt=%f"%global_dt
cwd=os.getcwd()
os.chdir(save_to_path)
readme = '''
## Model: %(CellID)s
### Original neuroConstruct config ID: %(SpikeProfile)s
**%(AnalysisHeader2)s**
s.png)
### Original neuroConstruct config ID: %(Config)s
**%(AnalysisHeader1)s**
s.png)
**IF curve for the NeuroML2 model simulated in NEURON**
s.png)
**IV curve for the NeuroML2 model simulated in NEURON**
s.png)
**Spike times versus time step: the NeuroML2 model simulated in NEURON.
Dashed black lines - spike times at the %(Smallest_dt)s ms time step; Green - spike times at the following time steps (in ms): %(DtArray)s.**
s.png)
**Spike times versus spatial discretization: the NeuroML2 model simulated in NEURON.
Default value for the number of internal divs is %(default_divs)s.
Dashed black lines - spike times at the %(MaximumDivs)s internal divisions; Blue - spike times at the following values of internal divisions:
%(IntDivsArray)s.**
s.png)'''
readme_file = open('README.md','w')
readme_final=readme%{"CellID":cellModel,
"IFcurve":IFcurve,
"IVcurve":IVcurve,
"Config":config_array[cellModel]['Analysis'],
"DtCurve":dt_curve,
"DxCurve":dx_curve,
"nC_vs_NML2Curve":analysis_string1,
"AnalysisHeader1":analysis_header1,
"SpikeProfileCurve":analysis_string2,
"AnalysisHeader2":analysis_header2,
"default_divs":default_num_of_comps,
"SpikeProfile":config_array[cellModel]['SpikeProfile'],
"Smallest_dt":smallest_dt,
"DtArray":dt_list,
"IntDivsArray":dx_array,
"MaximumDivs":maximum_int_divs}
readme_file.write(readme_final)
readme_file.close()
os.chdir(cwd)
else:
cwd=os.getcwd()
os.chdir(save_to_path)
readme = '''
## Model: %(CellID)s
### Original neuroConstruct config ID: %(Config)s
**%(AnalysisHeader1)s**
s.png)
**IF curve for the NeuroML2 model simulated in NEURON**
s.png)
**IV curve for the NeuroML2 model simulated in NEURON**
s.png)
**Spike times versus time step: the NeuroML2 model simulated in NEURON.
Dashed black lines - spike times at the %(Smallest_dt)s ms time step; Green - spike times for the following time steps (in ms): %(DtArray)s.**
s.png)
**Spike times versus spatial discretization: the NeuroML2 model simulated in NEURON.
Default value for the number of internal divs is %(default_divs)s.
Dashed black lines - spike times at the %(MaximumDivs)s internal divisions; Blue - spike times at the following values of internal divisions:
%(IntDivsArray)s.**
s.png)'''
readme_file = open('README.md','w')
readme_final=readme%{"CellID":cellModel,
"IFcurve":IFcurve,
"IVcurve":IVcurve,
"Config":config_array[cellModel]['Analysis'],
"DtCurve":dt_curve,
"DxCurve":dx_curve,
"nC_vs_NML2Curve":analysis_string1,
"AnalysisHeader1":analysis_header1,
"default_divs":default_num_of_comps,
"Smallest_dt":smallest_dt,
"DtArray":dt_list,
"IntDivsArray":dx_array,
"MaximumDivs":maximum_int_divs}
readme_file.write(readme_final)
readme_file.close()
os.chdir(cwd)
cwd=os.getcwd()
os.chdir(os.path.dirname(cwd))
readme = '''
## Conversion of Thalamocortical cell models to NeuroML2
'''
readme_file = open('README.md','w')
for cell_index in range(0,len(cell_id_list)):
readme_cell='''
## Model: %(CellID)s
### Original neuroConstruct config ID: %(Config)s
**%(AnalysisHeader)s**
s.png)
**IF curve for the NeuroML2 model simulated in NEURON**
s.png)'''
readme_cell=readme_cell%{"CellID":cell_id_list[cell_index],
"Config":config_list[cell_index],
"AnalysisHeader":analysis_header_list[cell_index],
"nC_vs_NML2Curve":os.path.join(cell_id_list[cell_index],nC_vs_NML2Curve_list[cell_index]),
"IFcurve":os.path.join(cell_id_list[cell_index],IFcurve_list[cell_index])}
readme=readme+readme_cell
readme_file.write(readme)
readme_file.close()
os.chdir(cwd)
from matplotlib import pyplot as plt
import sys
nml2_file = 'NML2_SingleCompHHCell.nml'
target = "net1"
dts = [0.1,0.05,0.01,0.005,0.001,0.0005,0.0001]
dts = [0.05,0.01,0.005]
nogui = '-nogui' in sys.argv # Used to supress GUI in tests for Travis-CI
analyse_spiketime_vs_dt(nml2_file,
target,
300,
'jNeuroML',
'hhpop[0]/v',
dts,
verbose=True,
show_plot_already=False)
if not nogui:
analyse_spiketime_vs_dt(nml2_file,
target,
300,
'jNeuroML_NEURON',
'hhpop[0]/v',
dts,
verbose=True,
show_plot_already=False)
if not nogui:
plt.show()
def dashboard_cells(net_id,
net_file_name,
config_array,
global_dt,
if_params,
elec_len_list,
dt_list,
generate_dashboards=True,
compare_to_neuroConstruct=False,
regenerate_nml2=False,
show_plot_already=False,
proj_string_neuroConstruct=None,
shell=None,
nc_home=None):
################### check whether use of neuroConstruct python/java interface is needed ########################
if regenerate_nml2:
use_NeuroConstruct(compare_to_neuroConstruct=compare_to_neuroConstruct,
regenerate_nml2=regenerate_nml2,
proj_string=proj_string_neuroConstruct,
global_dt=global_dt,
config_array=config_array,
elec_len_list=elec_len_list,
shell=shell,
nc_home=nc_home)
else:
use_NeuroConstruct(compare_to_neuroConstruct=compare_to_neuroConstruct,
regenerate_nml2=regenerate_nml2,
proj_string=proj_string_neuroConstruct,
global_dt=global_dt,
config_array=config_array,
shell=shell,
nc_home=nc_home)
############################################################################################################
if generate_dashboards:
cell_id_list = []
config_list = []
analysis_header_list = []
nC_vs_NML2Curve_list = []
IFcurve_list = []
for cellModel in config_array.keys():
cell_id_list.append(cellModel)
config_list.append(config_array[cellModel]['Analysis'])
save_to_path = "../" + cellModel
if not os.path.exists(save_to_path):
print("Creating a new directory %s" % save_to_path)
os.makedirs(save_to_path)
else:
print("A directory %s already exists" % save_to_path)
pathToConfig = "../" + config_array[cellModel]['Analysis']
try:
with open(os.path.join(pathToConfig, "compSummary.json"),
'r') as f:
comp_info = json.load(f)
except IOError:
print "cannot open file %s" % os.path.join(
pathToConfig, "compSummary.json")
cell_morph_summary = comp_info[config_array[cellModel]['Analysis']]
src_files = os.listdir(pathToConfig)
num_dx_configs = 0
dx_array = []
found_all_compartmentalizations = False
dx_configs = {}
target_v = None
found_default = False
for file_name in src_files:
full_file_path = os.path.join(pathToConfig, file_name)
if (os.path.isdir(full_file_path)) and "_default" in file_name:
found_default = True
original_LEMS_target = os.path.join(
full_file_path, "LEMS_Target.xml")
###################################################################################
if -1 in elec_len_list and "-1" in cell_morph_summary.keys(
):
dx_configs[cell_morph_summary["-1"]
["IntDivs"]] = original_LEMS_target
default_num_of_comps = cell_morph_summary["-1"][
"IntDivs"]
dx_array.append(int(default_num_of_comps))
num_dx_configs += 1
##################################################################################
print("%s is a directory" % full_file_path)
print("will generate the IF curve for %s.cell.nml" %
cellModel)
generate_current_vs_frequency_curve(
os.path.join(full_file_path, cellModel + ".cell.nml"),
cellModel,
start_amp_nA=if_params['start_amp_nA'],
end_amp_nA=if_params['end_amp_nA'],
step_nA=if_params['step_nA'],
analysis_duration=if_params['analysis_duration'],
analysis_delay=if_params['analysis_delay'],
dt=if_params['dt'],
temperature=if_params['temperature'],
plot_voltage_traces=if_params['plot_voltage_traces'],
plot_if=if_params['plot_if'],
plot_iv=if_params['plot_iv'],
show_plot_already=show_plot_already,
save_if_figure_to='%s/IF_%s.png' %
(save_to_path, cellModel),
save_iv_figure_to='%s/IV_%s.png' %
(save_to_path, cellModel),
simulator=if_params['simulator'])
IFcurve = "IF_%s" % cellModel
IFcurve_list.append(IFcurve)
IVcurve = "IV_%s" % cellModel
nml2_file_path = os.path.join(full_file_path,
net_file_name + ".net.nml")
net_doc = pynml.read_neuroml2_file(nml2_file_path)
net = net_doc.networks[0]
pop = net.populations[0]
popID = pop.id
target_v = "%s/0/%s/v" % (popID, cellModel)
########################################################################################
if if_params['simulator'] == 'jNeuroML':
results = pynml.run_lems_with_jneuroml(
original_LEMS_target,
nogui=True,
load_saved_data=True,
plot=False,
verbose=False)
if if_params['simulator'] == 'jNeuroML_NEURON':
results = pynml.run_lems_with_jneuroml_neuron(
original_LEMS_target,
nogui=True,
load_saved_data=True,
plot=False,
verbose=False)
t = results['t']
v = results[target_v]
if compare_to_neuroConstruct:
print(
"will generate the comparison between the nC model and NeuroML2 model"
)
PlotNC_vs_NML2(
{
'NML2': [{
't': t,
'v': v
}],
'nC': [config_array[cellModel]['OriginalTag']],
'subplotTitles': [
'NeuroML2 versus nC model: simulations in NEURON with dt=%f'
% global_dt
]
}, {
'cols': 8,
'rows': 5
},
legend=True,
show=False,
save_to_file='%s/nC_vs_NML2_%s.png' %
(save_to_path,
config_array[cellModel]['Analysis']),
nCcellPath=os.path.join(
save_to_path,
config_array[cellModel]['Analysis']))
analysis_string1 = "nC_vs_NML2_%s" % config_array[
cellModel]['Analysis']
analysis_header1 = "Comparison between the original nC model and NeuroML2 model: simulations in NEURON with dt=%f" % global_dt
analysis_header_list.append(analysis_header1)
nC_vs_NML2Curve_list.append(analysis_string1)
else:
print("will generate the plot for the NeuroML2 model")
generate_nml2_plot(
{
'NML2': [{
't': t,
'v': v
}],
'subplotTitles': [
'NeuroML2 model: simulations in NEURON with dt=%f'
% global_dt
]
}, {
'cols': 8,
'rows': 5
},
show=False,
save_to_file='%s/NML2_%s.png' %
(save_to_path,
config_array[cellModel]['Analysis']))
analysis_string1 = "NML2_%s" % config_array[cellModel][
'Analysis']
analysis_header1 = "NeuroML2 model: simulations in NEURON with dt=%f" % global_dt
analysis_header_array.append(analysis_header1)
nC_vs_NML2Curve_array.append(analysis_string1)
smallest_dt = min(dt_list)
########################################################################################
print(
"will generate the spike times vs dt curve for %s.cell.nml"
% cellModel)
analyse_spiketime_vs_dt(
nml2_file_path,
net_id,
get_sim_duration(
os.path.join(full_file_path,
"LEMS_%s.xml" % net_id)),
if_params['simulator'],
target_v,
dt_list,
verbose=False,
spike_threshold_mV=0,
show_plot_already=show_plot_already,
save_figure_to="%s/Dt_%s.png" %
(save_to_path, cellModel),
num_of_last_spikes=None)
dt_curve = "Dt_%s" % cellModel
if not found_all_compartmentalizations:
for elecLen in range(0, len(elec_len_list)):
elec_len = str(elec_len_list[elecLen])
if elec_len != "-1":
if (elec_len in file_name) and (
elec_len in cell_morph_summary.keys()):
dx_configs[cell_morph_summary[elec_len]
["IntDivs"]] = os.path.join(
full_file_path,
"LEMS_Target.xml")
num_dx_configs += 1
dx_array.append(
int(cell_morph_summary[elec_len]
["IntDivs"]))
break
if num_dx_configs == len(elec_len_list):
found_all_compartmentalizations = True
if not found_default:
print(
"default configuration for %s analysis is not found; execution will terminate; set regenerate_nml2 to True to generate target dirs."
% cellModel)
quit()
if found_all_compartmentalizations:
dx_array = list(np.sort(dx_array))
maximum_int_divs = max(dx_array)
print(
"testing the presence of cell configs with different levels of spatial discretization"
)
analyse_spiketime_vs_dx(dx_configs,
if_params['simulator'],
target_v,
verbose=False,
spike_threshold_mV=0,
show_plot_already=show_plot_already,
save_figure_to="%s/Dx_%s.png" %
(save_to_path, cellModel),
num_of_last_spikes=3)
dx_curve = "Dx_%s" % cellModel
else:
print(
"not all of the target configurations were recompartmentalized; execution will terminate; set regenerate_nml2 to True to obtain all of the target configurations."
)
quit()
if config_array[cellModel]['Analysis'] != config_array[cellModel][
'SpikeProfile']:
pathToProfileConfig = "../" + config_array[cellModel][
'SpikeProfile'] + "/" + config_array[cellModel][
'SpikeProfile'] + "_default"
original_LEMS_target = os.path.join(pathToProfileConfig,
"LEMS_Target.xml")
if if_params['simulator'] == 'jNeuroML':
results = pynml.run_lems_with_jneuroml(
original_LEMS_target,
nogui=True,
load_saved_data=True,
plot=False,
verbose=False)
if if_params['simulator'] == 'jNeuroML_NEURON':
results = pynml.run_lems_with_jneuroml_neuron(
original_LEMS_target,
nogui=True,
load_saved_data=True,
plot=False,
verbose=False)
if 'SpikeProfileTag' in config_array[cellModel].keys():
tag = config_array[cellModel][
'SpikeProfileTag'] + "_0_wtime"
else:
tag = config_array[cellModel]['OriginalTag']
if 'SpikeProfileCellTag' in config_array[cellModel].keys(
) and 'SpikeProfileTag' in config_array[cellModel].keys():
target_v = "%s/0/%s/v" % (
config_array[cellModel]['SpikeProfileTag'],
config_array[cellModel]['SpikeProfileCellTag'])
t = results['t']
v = results[target_v]
if compare_to_neuroConstruct:
print(
"will generate the comparison between the nC model and NeuroML2 model"
)
PlotNC_vs_NML2(
{
'NML2': [{
't': t,
'v': v
}],
'nC': [tag],
'subplotTitles': [
'NML2 versus nC model: simulations in NEURON with dt=%f'
% global_dt
]
}, {
'cols': 8,
'rows': 5
},
legend=True,
show=show_plot_already,
save_to_file='%s/nC_vs_NML2_%s.png' %
(save_to_path,
config_array[cellModel]['SpikeProfile']),
nCcellPath=os.path.join(
save_to_path,
config_array[cellModel]['SpikeProfile']))
analysis_string2 = "nC_vs_NML2_%s" % config_array[
cellModel]['SpikeProfile']
analysis_header2 = "Comparison between the original nC model and NeuroML2 model: simulations in NEURON with dt=%f" % global_dt
else:
print("will generate the plot for the NeuroML2 model")
generate_nml2_plot(
{
'NML2': [{
't': t,
'v': v
}],
'subplotTitles': [
'NeuroML2 model: simulations in NEURON with dt=%f'
% global_dt
]
}, {
'cols': 8,
'rows': 5
},
show=False,
save_to_file='%s/NML2_%s.png' %
(save_to_path,
config_array[cellModel]['SpikeProfile']))
analysis_string2 = "NML2_%s" % config_array[cellModel][
'SpikeProfile']
analysis_header2 = "NeuroML2 model: simulations in NEURON with dt=%f" % global_dt
cwd = os.getcwd()
os.chdir(save_to_path)
readme = '''
## Model: %(CellID)s
### Original neuroConstruct config ID: %(SpikeProfile)s
**%(AnalysisHeader2)s**
s.png)
### Original neuroConstruct config ID: %(Config)s
**%(AnalysisHeader1)s**
s.png)
**IF curve for the NeuroML2 model simulated in NEURON**
s.png)
**IV curve for the NeuroML2 model simulated in NEURON**
s.png)
**Spike times versus time step: the NeuroML2 model simulated in NEURON.
Dashed black lines - spike times at the %(Smallest_dt)s ms time step; Green - spike times at the following time steps (in ms): %(DtArray)s.**
s.png)
**Spike times versus spatial discretization: the NeuroML2 model simulated in NEURON.
Default value for the number of internal divs is %(default_divs)s.
Dashed black lines - spike times at the %(MaximumDivs)s internal divisions; Blue - spike times at the following values of internal divisions:
%(IntDivsArray)s.**
s.png)'''
readme_file = open('README.md', 'w')
readme_final = readme % {
"CellID": cellModel,
"IFcurve": IFcurve,
"IVcurve": IVcurve,
"Config": config_array[cellModel]['Analysis'],
"DtCurve": dt_curve,
"DxCurve": dx_curve,
"nC_vs_NML2Curve": analysis_string1,
"AnalysisHeader1": analysis_header1,
"SpikeProfileCurve": analysis_string2,
"AnalysisHeader2": analysis_header2,
"default_divs": default_num_of_comps,
"SpikeProfile": config_array[cellModel]['SpikeProfile'],
"Smallest_dt": smallest_dt,
"DtArray": dt_list,
"IntDivsArray": dx_array,
"MaximumDivs": maximum_int_divs
}
readme_file.write(readme_final)
readme_file.close()
os.chdir(cwd)
else:
cwd = os.getcwd()
os.chdir(save_to_path)
readme = '''
## Model: %(CellID)s
### Original neuroConstruct config ID: %(Config)s
**%(AnalysisHeader1)s**
s.png)
**IF curve for the NeuroML2 model simulated in NEURON**
s.png)
**IV curve for the NeuroML2 model simulated in NEURON**
s.png)
**Spike times versus time step: the NeuroML2 model simulated in NEURON.
Dashed black lines - spike times at the %(Smallest_dt)s ms time step; Green - spike times for the following time steps (in ms): %(DtArray)s.**
s.png)
**Spike times versus spatial discretization: the NeuroML2 model simulated in NEURON.
Default value for the number of internal divs is %(default_divs)s.
Dashed black lines - spike times at the %(MaximumDivs)s internal divisions; Blue - spike times at the following values of internal divisions:
%(IntDivsArray)s.**
s.png)'''
readme_file = open('README.md', 'w')
readme_final = readme % {
"CellID": cellModel,
"IFcurve": IFcurve,
"IVcurve": IVcurve,
"Config": config_array[cellModel]['Analysis'],
"DtCurve": dt_curve,
"DxCurve": dx_curve,
"nC_vs_NML2Curve": analysis_string1,
"AnalysisHeader1": analysis_header1,
"default_divs": default_num_of_comps,
"Smallest_dt": smallest_dt,
"DtArray": dt_list,
"IntDivsArray": dx_array,
"MaximumDivs": maximum_int_divs
}
readme_file.write(readme_final)
readme_file.close()
os.chdir(cwd)
cwd = os.getcwd()
os.chdir(os.path.dirname(cwd))
readme = '''
## Conversion of Thalamocortical cell models to NeuroML2
'''
readme_file = open('README.md', 'w')
for cell_index in range(0, len(cell_id_list)):
readme_cell = '''
## Model: %(CellID)s
### Original neuroConstruct config ID: %(Config)s
**%(AnalysisHeader)s**
s.png)
**IF curve for the NeuroML2 model simulated in NEURON**
s.png)'''
readme_cell = readme_cell % {
"CellID":
cell_id_list[cell_index],
"Config":
config_list[cell_index],
"AnalysisHeader":
analysis_header_list[cell_index],
"nC_vs_NML2Curve":
os.path.join(cell_id_list[cell_index],
nC_vs_NML2Curve_list[cell_index]),
"IFcurve":
os.path.join(cell_id_list[cell_index],
IFcurve_list[cell_index])
}
readme = readme + readme_cell
readme_file.write(readme)
readme_file.close()
os.chdir(cwd)
