def GeneticAlgorithm(sid,ignore_result=False):
desdata = DesignModel.objects.get(SID=sid)
subject = "NeuroDesign: optimisation process started"
sender = "NeuroDesign"
sendermail = "*****@*****.**"
message = "Your design optimisation has now started. You can follow the progress here:"+" http://development.neuropowertools.org/design/runGA/?retrieve="+str(desdata.shareID)+". Thank you for using NeuroDesign."
recipient = str(desdata.email)
key = settings.MAILGUN_KEY
command = "curl -s --user '" + key + "' https://api.mailgun.net/v3/neuropowertools.org/messages -F from='" + sender + \
" <" + sendermail + ">' -F to=" + recipient + " -F subject="+subject+" -F text='" + message + "'"
os.system(command)
matrices = probs_and_cons(sid)
if desdata.ITImodel == 1:
model = "fixed"
ITImean = desdata.ITIfixed
ITImin = None
ITImax = None
elif desdata.ITImodel == 2:
model = "exponential"
ITImean = desdata.ITItruncmean
ITImin = desdata.ITItruncmin
ITImax = desdata.ITItruncmax
elif desdata.ITImodel == 3:
model = "uniform"
ITImin = desdata.ITIunifmin
ITImax = desdata.ITIunifmax
ITImean = (desdata.ITIunifmin+desdata.ITIunifmax)/2.
EXP = geneticalgorithm.experiment(
TR = desdata.TR,
n_trials = desdata.L,
P = matrices['P'],
C = matrices['C'],
duration = desdata.duration,
n_stimuli = desdata.S,
rho = desdata.rho,
resolution = desdata.resolution,
stim_duration = desdata.stim_duration,
restnum = desdata.RestNum,
restdur = desdata.RestDur,
ITImodel = model,
ITImin = ITImin,
ITImean = ITImean,
ITImax = ITImax,
maxrep = desdata.MaxRepeat,
hardprob = desdata.HardProb,
t_prestim = desdata.t_prestim,
t_poststim = desdata.t_poststim
)
seed = np.random.randint(10000)
POP = geneticalgorithm.population(
experiment = EXP,
confoundorder = desdata.ConfoundOrder,
G = desdata.G,
R = [0.4,0.4,0.2],
q = desdata.q,
weights = desdata.W,
I = desdata.I,
preruncycles = desdata.preruncycles,
cycles = desdata.cycles,
convergence=desdata.conv_crit,
folder=desdata.onsetsfolder,
Aoptimality = True if desdata.Aoptimality == 1 else False,
seed=seed
)
POP.print_cmd()
desdata = DesignModel.objects.get(SID=sid)
runform = DesignRunForm(None, instance=desdata)
form = runform.save(commit=False)
form.running = 1
form.seed = seed
form.cmd = POP.cmd
form.save()
POP.naturalselection()
POP.download()
# Select optimal design
desdata = DesignModel.objects.get(SID=sid)
runform = DesignRunForm(None, instance=desdata)
form = runform.save(commit=False)
form.convergence = POP.finished
form.zip_filename = POP.zip_filename
form.zipfile = POP.file
form.save()
subject = "NeuroDesign: optimisation process ended"
sender = "NeuroDesign"
sendermail = "*****@*****.**"
message = "Your design optimisation has now ended. You can download the results here:"+" http://development.neuropowertools.org/design/runGA/?retrieve="+str(desdata.SID)+". Thank you for using NeuroDesign."
recipient = str(desdata.email)
key = settings.MAILGUN_KEY
command = "curl -s --user '" + key + "' https://api.mailgun.net/v3/neuropowertools.org/messages -F from='" + sender + \
" <" + sendermail + ">' -F to=" + recipient + " -F subject="+subject+" -F text='" + message + "'"
os.system(command)
from neurodesign import geneticalgorithm, generate, msequence, report
EXP = geneticalgorithm.experiment(TR=2,
n_trials=100,
P=[0.33, 0.33, 0.33],
C=[[1, 0, 0], [0, 1, 0], [0, 0, 1],
[1, -1, 0], [0, 1, -1]],
n_stimuli=3,
rho=0.3,
resolution=1,
stim_duration=1,
t_pre=0,
t_post=2,
restnum=0,
restdur=0,
ITImodel="exponential",
ITImin=1,
ITImean=2,
ITImax=4)
POP = geneticalgorithm.population(
experiment=EXP,
weights=[0, 0.5, 0.25, 0.25],
preruncycles=20,
cycles=20,
seed=1,
outdes=5,
folder=
'/Users/Joke/Documents/Onderzoek/ProjectsOngoing/Neuropower/playground/design'
)
EXP = geneticalgorithm.experiment(
TR=1,
duration=360,
P=[0.1667, 0.1667, 0.1667, 0.1667, 0.1667, 0.1667],
C=[
[1, 0, 0, 0, 0, 0],
[0, 1, 0, 0, 0, 0],
[0, 0, 1, 0, 0, 0],
[0, 0, 0, 1, 0, 0],
[0, 0, 0, 0, 1, 0],
[0, 0, 0, 0, 0, 1],
[1, -1, 0, 0, 0, 0],
[1, 0, -1, 0, 0, 0],
[1, 0, 0, -1, 0, 0],
[1, 0, 0, 0, -1, 0],
[1, 0, 0, 0, 0, -1],
[0, 1, -1, 0, 0, 0],
[0, 1, 0, -1, 0, 0],
[0, 1, 0, 0, -1, 0],
[0, 1, 0, 0, 0, -1],
[0, 0, 1, -1, 0, 0],
[0, 0, 1, 0, -1, 0],
[0, 0, 1, 0, 0, -1],
[0, 0, 0, 1, -1, 0],
[0, 0, 0, 1, 0, -1],
[0, 0, 0, 0, 1, -1],
],
n_stimuli=6,
rho=0.3,
resolution=0.1,
stim_duration=1,
t_pre=0,
t_post=2,
ITImodel="exponential",
ITImin=2,
ITImean=3,
ITImax=6)
from neurodesign import geneticalgorithm, generate, msequence
import numpy as np
# define experimental setup
EXP = geneticalgorithm.experiment(TR=2,
n_trials=20,
P=[0.3, 0.3, 0.4],
C=[[1, -1, 0], [0, 1, -1]],
n_stimuli=3,
rho=0.3,
stim_duration=1,
t_pre=0.5,
t_post=2,
ITImodel="uniform",
ITImin=2,
ITImax=4)
# define first design with a fixed ITI
DES = geneticalgorithm.design(
order=[0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1],
ITI=[2] * 20,
experiment=EXP)
# expand to design matrix
DES.designmatrix()
DES.FCalc(weights=[0, 0.5, 0.25, 0.25])
DES.FdCalc()
DES.FcCalc()
def GeneticAlgorithm(sid,ignore_result=False):
desdata = DesignModel.objects.filter(SID=sid).last()
matrices = probs_and_cons(sid)
if desdata.ITImodel == 1:
model = "fixed"
ITImean = desdata.ITIfixed
ITImin = None
ITImax = None
elif desdata.ITImodel == 2:
model = "exponential"
ITImean = desdata.ITItruncmean
ITImin = desdata.ITItruncmin
ITImax = desdata.ITItruncmax
elif desdata.ITImodel == 3:
model = "uniform"
ITImin = desdata.ITIunifmin
ITImax = desdata.ITIunifmax
ITImean = (desdata.ITIunifmin+desdata.ITIunifmax)/2.
if (desdata.MaxRepeat<4 and desdata.S<5) or desdata.S==2:
R = [0,1,0]
else:
R = [0.4,0.4,0.2]
print("R: "+str(R))
EXP = geneticalgorithm.experiment(
TR = desdata.TR,
n_trials = desdata.L,
P = matrices['P'],
C = matrices['C'],
duration = desdata.duration,
n_stimuli = desdata.S,
rho = desdata.rho,
resolution = desdata.resolution,
stim_duration = desdata.stim_duration,
restnum = desdata.RestNum,
restdur = desdata.RestDur,
ITImodel = model,
ITImin = ITImin,
ITImean = ITImean,
ITImax = ITImax,
confoundorder = desdata.ConfoundOrder,
maxrep = desdata.MaxRepeat,
hardprob = desdata.HardProb,
t_pre = desdata.t_prestim,
t_post = desdata.t_poststim
)
seed = np.random.randint(10000)
POP = geneticalgorithm.population(
experiment = EXP,
G = desdata.G,
R = R,
q = desdata.q,
weights = desdata.W,
I = desdata.I,
preruncycles = desdata.preruncycles,
cycles = desdata.cycles,
convergence=desdata.conv_crit,
folder=desdata.onsets_folder,
outdes=desdata.outdes,
Aoptimality = True if desdata.Aoptimality == 1 else False,
seed=seed
)
POP.print_cmd()
desdata = DesignModel.objects.filter(SID=sid).last()
runform = DesignRunForm(None, instance=desdata)
form = runform.save(commit=False)
form.taskstatus = 2
form.running = 1
form.seed = seed
form.cmd = POP.cmd
form.save()
local_naturalselection(POP,sid)
POP.download()
desdata = DesignModel.objects.filter(SID=sid).last()
runform = DesignRunForm(None, instance=desdata)
form = runform.save(commit=False)
form.finished = True
form.running = 0
form.taskstatus = 3
form.save()
# list all files (with full path) for report
infiles = [os.path.join(dp, f) for dp, dn, fn in os.walk(os.path.expanduser(desdata.onsets_folder)) for f in fn]
# strip away /var/tmp --> design_suffix/xxx
outfiles = [x[9:] for x in infiles]
for file in range(len(infiles)):
push_to_s3(infiles[file],"designs/"+outfiles[file])
# Select optimal design
desdata = DesignModel.objects.filter(SID=sid).last()
runform = DesignRunForm(None, instance=desdata)
form = runform.save(commit=False)
form.convergence = POP.finished
form.files = outfiles
form.save()
subject = "NeuroDesign: optimisation process ended"
sender = "NeuroDesign"
sendermail = "*****@*****.**"
message = "Your design optimisation has now ended. You can download the results here:"+" http://www.neuropowertools.org/design/runGA/?retrieve="+str(desdata.shareID)+". Thank you for using NeuroDesign."
recipient = str(desdata.email)
#recipient = '*****@*****.**'
key = settings.MAILGUN_KEY
command = "curl -s --user '" + key + "' https://api.mailgun.net/v3/neuropowertools.org/messages -F from='" + sender + \
" <" + sendermail + ">' -F to=" + recipient + " -F subject="+subject+" -F text='" + message + "'"
os.system(command)