desiredAngles[0] = utils.changeRange(
CX[t], #26.3,
CXHealthyRange[0],
CXHealthyRange[1],
shoulderRange[0],
shoulderRange[1])
desiredAngles[1] = utils.changeRange(CX[t], CXHealthyRange[0],
CXHealthyRange[1],
elbowRange[0], elbowRange[1])
Torque = arm.PD_controller([desiredAngles[0], desiredAngles[1]],
Kp1, Kp2, Kd1, Kd2) # compute torques
arm.SolveDirectDynamics(Torque[0], Torque[1]) # move the arm
efPosHistory[t, :] = np.array([arm.xEndEf, arm.yEndEf])
# if PLOTTING == True and t > start_plotting * 100:
# text1.set_text("sec = {}".format(t/100.0))
# xDesAng = arm.L1*np.cos(desiredAngles[0]) +\
# arm.L2*np.cos(desiredAngles[0]+desiredAngles[1])
# yDesAng = arm.L1*np.sin(desiredAngles[0]) +\
# arm.L2*np.sin(desiredAngles[0]+desiredAngles[1])
# desEnd.set_data([xDesAng],[yDesAng])
# armPlot.set_data([0,arm.xElbow, arm.xEndEf],
# [0,arm.yElbow, arm.yEndEf])
# plt.pause(dt)
efX = efPosHistory[:, 0].copy()
efY = efPosHistory[:, 1].copy()
class BrainDiffentialEquationModel:
def __init__(self,
modelType='DRNdamage',
configDir='bestConfigDir',
tStart=0,
tStop=200,
seedNum=1,
seedNoise=0.01,
dt=0.01,
dmgDRND7=True,
dmgDRND7Start1=50,
dmgDRND7Mag1=20,
dmgDRND7Start2=100,
dmgDRND7Start3=150,
dmgDRND7Mag2=30,
dmgDRND7Mag3=40,
saveData=True,
plotResults=True,
savePlot=True,
closePlot=True,
armPlot=True,
startArmPlot=0,
tremorPlot=True,
randomState=0):
"""
(self, modelType = 'oscillatory', configDir = 'bestConfigDir',
tStart = 0, tStop = 300, dt = 0.01, dmgSND8 = True, dmgSND8Start = 75,
dmgSND8Mag = +70., dmgDAOsc = -80, trmtD7 = True, trmtD7Start1 = 150,
trmtD7Mag1 = -10, trmtD7Start2 = 200, trmtD7Mag2 = -20,
trmtD7Start3 = 250, trmtD7Mag3 = -40, saveData = True,
plotResults = True, lines4Subplot = 2, closePlot = True,
armPlot = True, startArmPlot = 0, tremorPlot = True)
Simulate brain areas activities
Parameters
----------
modelType : TYPE str, optional
DESCRIPTION brain model type ('DRNdamage').
The default is 'DRNdamage'.
configDir : TYPE str, optional
DESCRITION set the configuration folder containing the model params
The default is 'bestConfigDir'.
tStart : TYPE int, optional
DESCRIPTION. simulation time start (s).
The default is 0.
tStop : TYPE int, optional
DESCRIPTION. simultion timne stop (s).
The default is 300.
seedNum : TYPE int, optional
DESCRIPTION set simulation seed number.
The default is 2
seedNoise : TYPE, float
DESCRIPTION set seed noise
The default is 0.0\
dt : TYPE float, optional
DESCRIPTION. simultion delta time(s).
The default is 0.01.
dmgDRND7 : TYPE bool, optional
DESCRIPTION. simulate SN damage (d8)).
The default is True.
dmgDRND7Start : TYPE int, optional
DESCRIPTION. SN damage start (s).
The default is 75.
dmgDRND7Mag1 : TYPE float, optional
DESCRIPTION. SN damage magnitude (%).
The default is +20.
dmgDRND7Mag2 : TYPE float, optional
DESCRIPTION. SN damage magnitude (%).
The default is +30.
dmgDRND7Mag3 : TYPE float, optional
DESCRIPTION. SN damage magnitude (%).
The default is +40.
saveData : TYPE bool, optional
DESCRIPTION. save simulation results.
The default is True.
plotResults : TYPE bool, optional
DESCRIPTION. plot brain areas activities.
The default is True.
lines4Subplot : TYPE int, optional
DESCRIPTION. Set number of line for subplot.
The default is 2.
savePlot : TYPE bool, optional
DESCRIPTION. Save simulation plot.
The default is True.
closePlot : TYPE bool, optional
DESCRIPTION. Automatically close simulation plot.
The default is True.
armPlot : TYPE bool, optional
DESCRIPTION. Plot arm simulation.
The default is True.
startArmPlot : TYPE int, optional
DESCRIPTION. Arm plot start plotting simulation time.
The default is 0.
tremorPlot : TYPE bool, optional
DESCRIPTION. Plot tremor amplitude results.
The default is true.
Returns
-------
None.
"""
self.modelType = modelType
self.configDir = configDir
self.tStart = tStart
self.tStop = tStop
self.seedNum = seedNum
self.seedNoise = seedNoise
self.dt = dt
self.dmgDRND7 = dmgDRND7
self.dmgDRND7Start1 = dmgDRND7Start1
self.dmgDRND7Start2 = dmgDRND7Start2
self.dmgDRND7Start3 = dmgDRND7Start3
self.dmgDRND7Mag1 = dmgDRND7Mag1
self.dmgDRND7Mag2 = dmgDRND7Mag2
self.dmgDRND7Mag3 = dmgDRND7Mag3
self.saveData = saveData
self.plotResults = plotResults
self.savePlot = savePlot
self.closePlot = closePlot
self.armPlot = armPlot
self.startArmPlot = startArmPlot
self.tremorPlot = tremorPlot
self.randomState = randomState
self.healtyRange = 25, 35
self.damagedRange1 = 75, 85
self.damagedRange2 = 125, 135
self.damagedRange3 = 175, 185
self.seedTremorData = np.zeros([self.seedNum, 4])
self.seedDA5HTData = np.zeros([self.seedNum, 4])
self.random = RandomState(randomState)
def makeModelDirs(self):
"""
Make brain model results directories
Returns
-------
None.
"""
self.cwd = os.getcwd()
# set data directory
self.resultsDir = os.path.join(self.cwd, 'results')
if not os.path.exists(self.resultsDir):
os.makedirs(self.resultsDir)
# set model directory
self.modelDir = os.path.join(self.resultsDir, self.modelType)
if not os.path.exists(self.modelDir):
os.makedirs(self.modelDir)
def getTimeParams(self):
"""
Get simulation time parameters
Returns
-------
None.
"""
self.nStep = int(self.tStop * (1 / self.dt) + 1)
self.time = np.linspace(self.tStart, self.tStop, self.nStep)
def getBrainModelParams(self):
"""
Get brain parameters for simulation
Parameters
----------
None.
Returns
-------
None.
"""
# init brain model parameters dictionary
self.brainModelParams = {}
# set configuration files directory
self.configDir = os.path.join(self.cwd, self.configDir)
# load brain areas steady state params
self.steadyStateParamsPath = os.path.join(
self.configDir, 'brainSteadyStateParams.json')
with open(self.steadyStateParamsPath) as fp:
self.steadyStateParams = json.load(fp)
self.brainModelParams.update(self.steadyStateParams)
#load model params
if self.modelType == 'DRNDamage':
self.oscillatoryParamsPath = os.path.join(
self.configDir, 'oscillatoryModelParams.json')
with open(self.oscillatoryParamsPath) as fp:
self.oscillatoryParams = json.load(fp)
self.brainModelParams.update(self.oscillatoryParams)
elif self.modelType == 'reed':
self.reedParamsPath = os.path.join(self.configDir,
'reedModelParams.json')
with open(self.reedParamsPath) as fp:
self.reedParams = json.load(fp)
self.brainModelParams.update(self.reedParams)
self.getModelSteadyStateParamsList()
def getModelSteadyStateParamsList(self):
"""
Get brain elements list from brain parameters
Returns
-------
None.
"""
self.modelSteadyStateParamsNameList = []
for key in list(self.brainModelParams.keys()):
if key[-2:] == '_0':
self.modelSteadyStateParamsNameList.append(key[:-2])
def getSeedParams(self):
"""
get seeds noised parameters
"""
self.seedsParamsList = []
self.modelParamValuesList = list(self.brainModelParams.values())
self.modelParamNamesList = list(self.brainModelParams.keys())
for i in range(self.seedNum):
seedParams = {}
for i, param in enumerate(self.modelParamNamesList):
seedParams[param] = \
self.modelParamValuesList[i] + self.random.normal(
0.0,
self.modelParamValuesList[i] * self.seedNoise
)
self.seedsParamsList.append(seedParams)
def makeSeedDir(self, seed):
self.seedDir = os.path.join(self.modelDir, 'seed_{}'.format(seed))
if not os.path.exists(self.seedDir):
os.makedirs(self.seedDir)
# set brain data directory
self.brainDir = os.path.join(self.seedDir, 'brainActivities')
if not os.path.exists(self.brainDir):
os.makedirs(self.brainDir)
# set plot directory
self.plotsDir = os.path.join(self.seedDir, 'plots')
if not os.path.exists(self.plotsDir):
os.makedirs(self.plotsDir)
# set arm data directory
self.armDir = os.path.join(self.seedDir, 'armSimulation')
if not os.path.exists(self.armDir):
os.makedirs(self.armDir)
def getBrainElementsInitialActivities(self):
"""
Get brain areas initial activities
Returns
-------
None.
"""
self.y0 = [
self.brainModelParams['IP_0'], self.brainModelParams['DP_0'],
self.brainModelParams['Thal_0'], self.brainModelParams['M1_0'],
self.brainModelParams['DRN_0'], self.brainModelParams['DA_0'],
self.brainModelParams['5-HT_0'], self.brainModelParams['SNc_0']
]
def integrateBrainElementsActivities(self):
"""
Integrate brain areas differential equation system
Returns
-------
None.
"""
if self.modelType == 'DRNDamage':
self.y = odeint(
drnDamageEquationsSystem, # differential equation system
self.y0, # initial state
self.time, # integration step
args=(
self.brainModelParams['a1c'],
self.brainModelParams['a1da'], self.brainModelParams['d1'],
self.brainModelParams['aOsc'],
self.brainModelParams['fOsc'],
self.brainModelParams['daOsc'],
self.brainModelParams['a2c'],
self.brainModelParams['a2da'], self.brainModelParams['d2'],
self.brainModelParams['a3'], self.brainModelParams['a3md'],
self.brainModelParams['a3mi'], self.brainModelParams['d3'],
self.brainModelParams['a4th'], self.brainModelParams['d4'],
self.brainModelParams['a5'], self.brainModelParams['a5cx'],
self.brainModelParams['a5sn'], self.brainModelParams['d5'],
self.brainModelParams['G'], self.brainModelParams['d6'],
self.brainModelParams['a7'], self.brainModelParams['d7'],
self.brainModelParams['a8'],
self.brainModelParams['a8drn'],
self.brainModelParams['d8'], self.dmgDRND7,
self.dmgDRND7Start1, self.dmgDRND7Mag1,
self.dmgDRND7Start2, self.dmgDRND7Mag2,
self.dmgDRND7Start3, self.dmgDRND7Mag3,
self.brainModelParams['d5'], self.brainModelParams['d7']),
)
def getPlotVLines(self):
"""
Get model plot labels
Returns
-------
None.
"""
self.vLinesLabelList = []
self.vLinesXList = []
# Get damaged model label
if self.dmgDRND7:
self.dmgLbl1 = '$\u03C4_{_{5HT}}$ = +20%'
self.vLinesLabelList.append(self.dmgLbl1)
self.vLinesXList.append(self.dmgDRND7Start1)
self.dmgLbl2 = '$\u03C4_{_{5HT}}$ = +30%'
self.vLinesLabelList.append(self.dmgLbl2)
self.vLinesXList.append(self.dmgDRND7Start2)
self.dmgLbl3 = '$\u03C4_{_{5HT}}$ = +40%'
self.vLinesLabelList.append(self.dmgLbl3)
self.vLinesXList.append(self.dmgDRND7Start3)
def plotBrainElementsActivities(self):
"""
Plot brain elements activities
Returns
-------
None.
"""
self.getPlotVLines()
plot1Name = self.modelSteadyStateParamsNameList[5] + '-' + \
self.modelSteadyStateParamsNameList[6] + ' ' + \
'activities'
yLabel = 'Concentration [nM]'
# funcList = [self.y[:,0],self.y[:,1]]
# labelList = [self.modelSteadyStateParamsNameList[0],
# self.modelSteadyStateParamsNameList[1]]
savePlotPath = os.path.join(self.plotsDir, plot1Name + '.png')
plotFigure = plt.figure(plot1Name, dpi=300)
plotAxes = plotFigure.add_subplot(111)
plotAxes.set_xlabel('Time [sec]', fontsize=50, fontweight='bold')
plotAxes.set_ylabel(yLabel, fontsize=50, fontweight='bold')
plotAxes.set_xlim(self.time[0], self.time[-1])
plotAxes.set_ylim(0, 5)
plt.xticks(fontsize=50)
plt.yticks(fontsize=50)
vline1 = plotAxes.axvline(x=self.vLinesXList[0],
ls='--',
label=self.dmgLbl1,
color='black',
linewidth=10)
vline2 = plotAxes.axvline(x=self.vLinesXList[1],
ls='--',
label=self.dmgLbl2,
color='lightgrey',
linewidth=10)
vline3 = plotAxes.axvline(x=self.vLinesXList[2],
ls='--',
label=self.dmgLbl3,
color='grey',
linewidth=10)
line1, = plotAxes.plot(self.time,
self.y[:, 5],
label=self.modelSteadyStateParamsNameList[5],
linewidth=10,
color='pink')
line2, = plotAxes.plot(self.time,
self.y[:, 6],
label=self.modelSteadyStateParamsNameList[6],
linewidth=10,
color='brown')
plotAxes.legend((
line1,
line2,
vline1,
vline2,
vline3,
), (
self.modelSteadyStateParamsNameList[5],
self.modelSteadyStateParamsNameList[6],
self.dmgLbl1,
self.dmgLbl2,
self.dmgLbl3,
),
loc='best',
fontsize=25)
plotFigure.savefig(savePlotPath)
if self.closePlot:
plt.close()
plot1Name = self.modelSteadyStateParamsNameList[4] + '-' + \
self.modelSteadyStateParamsNameList[7] + ' ' + \
'activities'
yLabel = 'Spyke frequency [Hz]'
# funcList = [self.y[:,0],self.y[:,1]]
# labelList = [self.modelSteadyStateParamsNameList[0],
# self.modelSteadyStateParamsNameList[1]]
savePlotPath = os.path.join(self.plotsDir, plot1Name + '.png')
plotFigure = plt.figure(plot1Name, figsize=(32, 16))
plotAxes = plotFigure.add_subplot(111)
plotAxes.set_xlabel('Time [sec]', fontsize=50, fontweight='bold')
plotAxes.set_ylabel(yLabel, fontsize=50, fontweight='bold')
plotAxes.set_xlim(self.time[0], self.time[-1])
plotAxes.set_ylim(0, 5)
plt.xticks(fontsize=50)
plt.yticks(fontsize=50)
vline1 = plotAxes.axvline(x=self.vLinesXList[0],
ls='--',
label=self.dmgLbl1,
color='black',
linewidth=10)
vline2 = plotAxes.axvline(x=self.vLinesXList[1],
ls='--',
label=self.dmgLbl2,
color='lightgrey',
linewidth=10)
vline3 = plotAxes.axvline(x=self.vLinesXList[2],
ls='--',
label=self.dmgLbl3,
color='grey',
linewidth=10)
line1, = plotAxes.plot(self.time,
self.y[:, 4],
label=self.modelSteadyStateParamsNameList[4],
linewidth=10,
color='red')
line2, = plotAxes.plot(self.time,
self.y[:, 7],
label=self.modelSteadyStateParamsNameList[7],
linewidth=10,
color='blue')
plotAxes.legend((
line1,
line2,
vline1,
vline2,
vline3,
), (
self.modelSteadyStateParamsNameList[4],
self.modelSteadyStateParamsNameList[7],
self.dmgLbl1,
self.dmgLbl2,
self.dmgLbl3,
),
loc='best',
fontsize=25)
plotFigure.savefig(savePlotPath)
if self.closePlot:
plt.close()
plot1Name = self.modelSteadyStateParamsNameList[2] + '-' + \
self.modelSteadyStateParamsNameList[3] + ' ' + \
'activities'
yLabel = 'Spyke frequency [Hz]'
# funcList = [self.y[:,0],self.y[:,1]]
# labelList = [self.modelSteadyStateParamsNameList[0],
# self.modelSteadyStateParamsNameList[1]]
savePlotPath = os.path.join(self.plotsDir, plot1Name + '.png')
plotFigure = plt.figure(plot1Name, figsize=(32, 16))
plotAxes = plotFigure.add_subplot(111)
plotAxes.set_xlabel('Time [sec]', fontsize=50, fontweight='bold')
plotAxes.set_ylabel(yLabel, fontsize=50, fontweight='bold')
plotAxes.set_xlim(self.time[0], self.time[-1])
plotAxes.set_ylim(0, 35)
plt.xticks(fontsize=50)
plt.yticks(fontsize=50)
vline1 = plotAxes.axvline(x=self.vLinesXList[0],
ls='--',
label=self.dmgLbl1,
color='black',
linewidth=10)
vline2 = plotAxes.axvline(x=self.vLinesXList[1],
ls='--',
label=self.dmgLbl2,
color='lightgrey',
linewidth=10)
vline3 = plotAxes.axvline(x=self.vLinesXList[2],
ls='--',
label=self.dmgLbl3,
color='grey',
linewidth=10)
line1, = plotAxes.plot(self.time,
self.y[:, 2],
label=self.modelSteadyStateParamsNameList[2],
linewidth=10,
color='purple')
line2, = plotAxes.plot(self.time,
self.y[:, 3],
label=self.modelSteadyStateParamsNameList[3],
linewidth=10,
color='orange')
plotAxes.legend((
line1,
line2,
vline1,
vline2,
vline3,
), (
self.modelSteadyStateParamsNameList[2],
self.modelSteadyStateParamsNameList[3],
self.dmgLbl1,
self.dmgLbl2,
self.dmgLbl3,
),
loc='best',
fontsize=25)
plotFigure.savefig(savePlotPath)
if self.closePlot:
plt.close()
plot1Name = self.modelSteadyStateParamsNameList[0] + '-' + \
self.modelSteadyStateParamsNameList[1] + ' ' + \
'activities'
yLabel = 'Spyke frequency [Hz]'
# funcList = [self.y[:,0],self.y[:,1]]
# labelList = [self.modelSteadyStateParamsNameList[0],
# self.modelSteadyStateParamsNameList[1]]
savePlotPath = os.path.join(self.plotsDir, plot1Name + '.png')
plotFigure = plt.figure(plot1Name, figsize=(32, 16))
plotAxes = plotFigure.add_subplot(111)
plotAxes.set_xlabel('Time [sec]', fontsize=50, fontweight='bold')
plotAxes.set_ylabel(yLabel, fontsize=50, fontweight='bold')
plotAxes.set_xlim(self.time[0], self.time[-1])
plotAxes.set_ylim(0, 5)
plt.xticks(fontsize=50)
plt.yticks(fontsize=50)
vline1 = plotAxes.axvline(x=self.vLinesXList[0],
ls='--',
label=self.dmgLbl1,
color='black',
linewidth=10)
vline2 = plotAxes.axvline(x=self.vLinesXList[1],
ls='--',
label=self.dmgLbl2,
color='lightgrey',
linewidth=10)
vline3 = plotAxes.axvline(x=self.vLinesXList[2],
ls='--',
label=self.dmgLbl3,
color='grey',
linewidth=10)
line1, = plotAxes.plot(self.time,
self.y[:, 0],
label=self.modelSteadyStateParamsNameList[0],
linewidth=10,
color='lightgreen')
line2, = plotAxes.plot(self.time,
self.y[:, 1],
label=self.modelSteadyStateParamsNameList[1],
linewidth=10,
color='green')
plotAxes.legend((
line1,
line2,
vline1,
vline2,
vline3,
), (
self.modelSteadyStateParamsNameList[0],
self.modelSteadyStateParamsNameList[1],
self.dmgLbl1,
self.dmgLbl2,
self.dmgLbl3,
),
loc='best',
fontsize=25)
plotFigure.savefig(savePlotPath)
if self.closePlot:
plt.close()
def getTremorOscillationAmplitude(self, intervalRange):
"""
compute tremor oscillation amplitude
"""
timeRange = [intervalRange[0] * 100, intervalRange[1] * 100]
tremDist = self.euclDist[timeRange[0]:timeRange[1]]
dTremDist = np.ediff1d(tremDist, to_begin=np.array([0]))
signDTremDist = np.sign(dTremDist)
diffSignDTremDist = np.ediff1d(signDTremDist)
tremDistMin = np.where(diffSignDTremDist > 1)[0]
oscillationAmplitudeTrem = np.array([])
# oscillationAmplitudeTremSum = 0
for i in range(len(tremDistMin) - 1):
oscillationAmplitudeTremSum = 0
oscillationIntervalTrem = \
tremDist[tremDistMin[i]:tremDistMin[i+1]]
oscillationAmplitudeTremSum += \
np.sum(oscillationIntervalTrem)
oscillationAmplitudeTrem = \
np.hstack([oscillationAmplitudeTrem,
oscillationAmplitudeTremSum])
return np.mean(oscillationAmplitudeTrem), np.std(
oscillationAmplitudeTrem)
def saveEvents(self):
"""
Save simulation events data
Returns
-------
None.
"""
if self.modelType == 'DRNDamage':
self.eventsData = {
'd7 augmentation start 1': self.dmgDRND7Start1,
'd7 augmentation mag 1': self.dmgDRND7Mag2,
'd7 augmentation start 2': self.dmgDRND7Start2,
'd7 augmentation mag 2': self.dmgDRND7Mag2,
'd7 augmentation start 3': self.dmgDRND7Start3,
'd7 augmentation mag 3': self.dmgDRND7Mag3,
}
self.eventsPath = os.path.join(self.brainDir, 'eventData.json')
with open(self.eventsPath, 'w') as fp:
json.dump(self.eventsData, fp)
def saveBrainElementsActivities(self):
"""
Save brain elements activities in csv format
Returns
-------
None.
"""
self.saveArray = np.hstack([self.time.reshape(-1, 1), self.y])
self.dataColNames = ['Time'] + self.modelSteadyStateParamsNameList
self.activitiesDataframe = pd.DataFrame(data=self.saveArray,
columns=self.dataColNames)
self.dataframeSavePath = os.path.join(self.brainDir,
'brainElementsActivities.csv')
self.activitiesDataframe.to_csv(self.dataframeSavePath)
def runSim(self):
self.makeModelDirs()
self.getTimeParams()
self.getBrainModelParams()
self.getSeedParams()
for seed, seedParams in enumerate(self.seedsParamsList):
print('\nSeed {}'.format(seed))
self.brainModelParams = copy.copy(seedParams)
self.makeSeedDir(seed)
self.getBrainElementsInitialActivities()
self.integrateBrainElementsActivities()
self.plotBrainElementsActivities()
# self.getHealtySteadyStateMAPE()
# if self.saveData:
# self.saveEvents()
# self.saveBrainElementsActivities()
self.CX = self.y[:, 3].copy()
self.CXHealthyMean = self.CX[self.healtyRange[0] *
100:self.healtyRange[1] * 100].mean()
self.CXHealthyRange = [0, self.CXHealthyMean * 2]
self.efPosHistory = np.zeros([self.nStep, 2])
self.DA = self.y[:, 5].copy()
self.DAHealthyMean = self.DA[self.healtyRange[0] *
100:self.healtyRange[1] * 100].mean()
self.DADamaged1Mean = self.DA[self.damagedRange1[0] *
100:self.damagedRange1[1] *
100].mean()
self.DADamaged2Mean = self.DA[self.damagedRange2[0] *
100:self.damagedRange2[1] *
100].mean()
self.DADamaged3Mean = self.DA[self.damagedRange3[0] *
100:self.damagedRange3[1] *
100].mean()
self.shoulderRange = np.deg2rad(np.array([-60.0, 150.0]))
self.elbowRange = np.deg2rad(np.array([0.0, 180.0]))
self.arm = Arm(self.shoulderRange, self.elbowRange)
self.desiredAngles = np.ones(2) * 0.5
self.Kp1 = 20.0
self.Kd1 = 1.5
self.Kp2 = 10.0
self.Kd2 = 1.0
if self.armPlot == True:
self.simFig = plt.figure("ARM SIMULATION", dpi=300)
self.simPlot = self.simFig.add_subplot(111)
self.simPlot.set_xlim([-1.0, 1.0])
self.simPlot.set_ylim([0.0, 1.0])
self.text1 = plt.figtext(.02,
.72,
"sec = %s" % (0.00),
style='italic',
bbox={'facecolor': 'yellow'})
self.armPlot, = self.simPlot.plot(
[0, self.arm.xElbow, self.arm.xEndEf],
[0, self.arm.yElbow, self.arm.yEndEf],
'k-',
color='black',
linewidth=5)
self.xDesAng = self.arm.L1*np.cos(self.desiredAngles[0]) + \
self.arm.L2*np.cos(self.desiredAngles[0]+self.desiredAngles[1])
self.yDesAng = self.arm.L1*np.sin(self.desiredAngles[0]) + \
self.arm.L2*np.sin(self.desiredAngles[0]+self.desiredAngles[1])
self.desEnd, = self.simPlot.plot([self.xDesAng],
[self.yDesAng],
'o',
color='green',
markersize=10)
for t in range(self.nStep):
self.desiredAngles[0] = utils.changeRange(
self.CX[t], self.CXHealthyRange[0], self.CXHealthyRange[1],
self.shoulderRange[0], self.shoulderRange[1])
self.desiredAngles[1] = utils.changeRange(
self.CX[t], self.CXHealthyRange[0], self.CXHealthyRange[1],
self.elbowRange[0], self.elbowRange[1])
self.Torque = self.arm.PD_controller(
[self.desiredAngles[0], self.desiredAngles[1]], self.Kp1,
self.Kp2, self.Kd1, self.Kd2) # compute torques
self.arm.SolveDirectDynamics(self.Torque[0],
self.Torque[1]) # move the arm
self.efPosHistory[t, :] = np.array(
[self.arm.xEndEf, self.arm.yEndEf])
if self.armPlot and t > self.startArmPlot:
self.text1.set_text("sec = {}".format(t / 100.0))
self.xDesAng = self.arm.L1*np.cos(self.desiredAngles[0]) +\
self.arm.L2*np.cos(self.desiredAngles[0]+self.desiredAngles[1])
self.yDesAng = self.arm.L1*np.sin(self.desiredAngles[0]) +\
self.arm.L2*np.sin(self.desiredAngles[0]+self.desiredAngles[1])
self.desEnd.set_data([self.xDesAng], [self.yDesAng])
self.armPlot.set_data(
[0, self.arm.xElbow, self.arm.xEndEf],
[0, self.arm.yElbow, self.arm.yEndEf])
plt.pause(self.dt)
if self.saveData:
efColNames = ['X', 'Y']
self.efDataFrame = pd.DataFrame(data=self.efPosHistory,
columns=efColNames)
efPath = os.path.join(self.armDir, 'efPos.csv')
self.efDataFrame.to_csv(efPath)
self.efX = self.efPosHistory[:, 0].copy()
self.efY = self.efPosHistory[:, 1].copy()
self.efdX = np.ediff1d(self.efX, to_begin=np.array([0]))
self.efdY = np.ediff1d(self.efY, to_begin=np.array([0]))
self.euclDist = np.hypot(self.efdX, self.efdY)
self.meanPhys, self.stdPhys = \
self.getTremorOscillationAmplitude(self.healtyRange)
if self.modelType == 'DRNDamage':
self.meanDmg1, self.stdDmg1 = \
self.getTremorOscillationAmplitude(self.damagedRange1)
self.meanDmg2, self.stdDmg2 = \
self.getTremorOscillationAmplitude(self.damagedRange2)
self.meanDmg3, self.stdDmg3 = \
self.getTremorOscillationAmplitude(self.damagedRange3)
self.seedTremorData[seed, 0] = self.meanPhys
if self.modelType == 'DRNDamage':
self.seedTremorData[seed, 1] = self.meanDmg1
self.seedTremorData[seed, 2] = self.meanDmg2
self.seedTremorData[seed, 3] = self.meanDmg3
self.seedDA5HTData[seed, 0] = self.DAHealthyMean
self.seedDA5HTData[seed, 1] = self.DADamaged1Mean
self.seedDA5HTData[seed, 2] = self.DADamaged2Mean
self.seedDA5HTData[seed, 3] = self.DADamaged3Mean
self.dataColumns = ['healthy', 'damaged 1', 'damaged 2', 'damaged 3']
self.tremorResults = pd.DataFrame(data=self.seedTremorData,
columns=self.dataColumns)
if self.tremorPlot:
self.healthyLabel = 'HEALTHY'
self.xTicksLabels = [
self.healthyLabel,
'DAMAGE 1',
'DAMAGE 2',
'DAMAGE 3',
]
self.oscillationAmpFig = \
plt.figure(
"oscillation amplitude",
dpi=300
)
self.oscillationAmpPlot = self.oscillationAmpFig.add_subplot(111)
self.oscillationAmpPlot.set_ylabel("Oscillation amplitude (m)",
fontsize=10,
fontweight='bold')
self.oscillationAmpPlot.set_xlabel("DRN Lesion",
fontsize=10,
fontweight='bold')
self.oscillationAmpPlot.set_ylim(0.0, 0.01)
self.oscillationAmpPlot.set_xticks(np.arange(5))
self.oscillationAmpPlot.set_xticklabels(self.xTicksLabels)
self.oscillationAmpPlot.tick_params(axis='both',
which='major',
labelsize=7)
self.oscillationAmpPlot.tick_params(axis='both',
which='minor',
labelsize=7)
self.oscillationAmpPlot.errorbar([0],
self.seedTremorData[:, 0].mean(),
yerr=self.seedTremorData[:,
0].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='solid')
self.oscillationAmpPlot.errorbar([1],
self.seedTremorData[:, 1].mean(),
yerr=self.seedTremorData[:,
1].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='solid')
self.oscillationAmpPlot.errorbar([2],
self.seedTremorData[:, 2].mean(),
yerr=self.seedTremorData[:,
2].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='solid')
self.oscillationAmpPlot.errorbar([3],
self.seedTremorData[:, 3].mean(),
yerr=self.seedTremorData[:,
3].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='solid')
self.oscillationAmpPlot.plot(list(range(4)), [
self.seedTremorData[:, 0].mean(),
self.seedTremorData[:, 1].mean(),
self.seedTremorData[:, 2].mean(),
self.seedTremorData[:, 3].mean()
],
color='black')
if self.saveData:
self.dataColumns = [
'healthy', 'damaged1', 'damaged2', 'damaged3'
]
self.tremorResults = pd.DataFrame(data=self.seedTremorData,
columns=self.dataColumns)
tremorPath = os.path.join(self.modelDir, 'tremors.csv')
self.tremorResults.to_csv(tremorPath)
self.DAConcFig = \
plt.figure(
"DA concentration reduction",
dpi=300
)
self.DAConcPlot = self.DAConcFig.add_subplot(111)
self.DAConcPlot.set_ylabel("DA concentration (nM)",
fontsize=10,
fontweight='bold')
self.DAConcPlot.set_xlabel("DRN Lesion",
fontsize=10,
fontweight='bold')
# self.DAConcPlot.set_ylim(0.0, 3.0)
self.DAConcPlot.set_xticks(np.arange(5))
self.DAConcPlot.set_xticklabels(self.xTicksLabels)
self.DAConcPlot.tick_params(
axis='both',
which='major',
labelsize=7,
# fontweight= 'bold'
)
self.DAConcPlot.tick_params(
axis='both',
which='minor',
labelsize=7,
# fontweight= 'bold'
)
self.DAConcPlot.errorbar([0],
self.seedDA5HTData[:, 0].mean(),
yerr=self.seedDA5HTData[:, 0].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='none')
self.DAConcPlot.errorbar([1],
self.seedDA5HTData[:, 1].mean(),
yerr=self.seedDA5HTData[:, 1].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='none')
self.DAConcPlot.errorbar([2],
self.seedDA5HTData[:, 2].mean(),
yerr=self.seedDA5HTData[:, 2].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='none')
self.DAConcPlot.errorbar([3],
self.seedDA5HTData[:, 3].mean(),
yerr=self.seedDA5HTData[:, 3].std(),
marker='s',
markersize='7',
mec='black',
mfc='white',
ecolor='black',
fmt='',
linestyle='none')
self.DAConcPlot.plot(list(range(4)), [
self.seedDA5HTData[:, 0].mean(), self.seedDA5HTData[:,
1].mean(),
self.seedDA5HTData[:, 2].mean(), self.seedDA5HTData[:,
3].mean()
],
color='black')
if self.saveData:
self.dataColumns = [
'healthy', 'damaged1', 'damaged2', 'damaged3'
]
self.DAResults = pd.DataFrame(data=self.seedDA5HTData,
columns=self.dataColumns)
DAPath = os.path.join(self.modelDir, 'DAconc.csv')
self.DAResults.to_csv(DAPath)
