def update(val):
freq = int(sfreq.val)
print(freq)
l.set_ydata(data.data[val])
l2.set_ydata(sig.medfilt(data.data[val], 5))
temp2 = sig.medfilt(data.data[val], 5)
temp = SIFdata._assembleInit(data.wavelength, data.data[val])
l3.set_ydata(_gauss.gauss(data.wavelength, temp[1:]) + temp[0])
l4.set_ydata(
SIFdata._fitData4(val,
data.wavelength,
data.data,
bounds=True,
tol=1e-6))
fig.canvas.draw_idle()
def plotout(xdata, ydata, inp):
#show the input vs output of the fit versus the data given by xdata, ydata and the inp
temp = SIFdata._assembleInit(xdata, ydata)
plt.plot(xdata, ydata)
plt.plot(xdata, _gauss.gauss(xdata, inp[1:]) + inp[0], 'c')
for i in xrange((len(inp) - 1) / 3):
plt.plot(xdata,
inp[0] + _gauss.gauss(xdata, inp[3 * i + 1:3 * (i + 1) + 1]),
'r:')
for i in xrange((len(temp) - 1) / 3):
plt.plot(xdata,
temp[0] + _gauss.gauss(xdata, temp[3 * i + 1:3 *
(i + 1) + 1]),
'b-.',
alpha=.7)
def plotSIF9(shot, init=0, offset=0):
#DATA GET
data = SIFdata.SIFData(shot, offset=offset)
fig, ax = plt.subplots()
plt.subplots_adjust(bottom=0.25)
#plt.axvline(ArHe_like_lambda['w'],color='b')
#plt.axvline(ArHe_like_lambda['x'],color='b')
#plt.axvline(ArHe_like_lambda['y'],color='b')
#plt.axvline(ArHe_like_lambda['z'],color='b')
#plt.axvline(ArLi_like_lambda['a'],color='r')
#plt.axvline(ArLi_like_lambda['j'],color='r')
#plt.axvline(ArLi_like_lambda['q'],color='r')
#plt.axvline(ArLi_like_lambda['r'],color='r')
#plt.axvline(ArLi_like_lambda['s'],color='r')
#plt.axvline(ArLi_like_lambda['t'],color='r')
plt.axvline(Wlambda['px375'], color='g')
temp = ['r', 'g', 'b', 'k', 'c', 'y']
inp = 0
temp2 = [['W', 43, 'c'], ['W', 44, 'c'], ['W', 45, 'c'], ['Ar', 16, 'b']]
for i in temp2:
print(i, temp[inp])
for j in SIFdata.lines[i[0]][i[1]]:
plt.axvline(j, color=i[2])
l, = plt.semilogy(data.wavelength, data.data[init], lw=2, color='green')
l2, = plt.semilogy(data.wavelength,
sig.medfilt(data.data[init], 5),
lw=2,
color='red')
temp = SIFdata._assembleInit(data.wavelength, data.data[init])
l3, = plt.semilogy(data.wavelength,
_gauss.gauss(data.wavelength, temp[1:]) + temp[0],
lw=2.,
color='c')
l4, = plt.semilogy(data.wavelength,
SIFdata._fitData4(init,
data.wavelength,
data.data,
bounds=True),
lw=2.,
color='m',
alpha=.5)
#l5, = plt.semilogy(data.wavelength, SIFdata._fitData9(init, data.wavelength, data.data, bounds=True), lw=2.,color='k',alpha=.5)
l6, = plt.semilogy(data.wavelength,
SIFdata._fitData8(init,
data.wavelength,
data.data,
bounds=True),
lw=2.,
color='b',
alpha=.5)
plt.ylim((1e3, 5e4)) #scipy.nanmax(data.data)))
axcolor = 'lightgoldenrodyellow'
axfreq = plt.axes([0.25, 0.1, 0.65, 0.03])
sfreq = DiscreteSlider(axfreq,
'Freq',
0,
len(data.time),
valinit=init,
valfmt='%0.0i',
increment=1.0)
def update(val):
freq = int(sfreq.val)
print(freq)
l.set_ydata(data.data[val])
l2.set_ydata(sig.medfilt(data.data[val], 5))
temp2 = sig.medfilt(data.data[val], 7)
temp = SIFdata._assembleInit(data.wavelength, data.data[val])
l3.set_ydata(_gauss.gauss(data.wavelength, temp[1:]) + temp[0])
l4.set_ydata(
SIFdata._fitData4(val,
data.wavelength,
data.data,
bounds=True,
tol=1e-8))
#l5.set_ydata(SIFdata._fitData9(val, data.wavelength, data.data, bounds=True,tol=1e-8))
l6.set_ydata(
SIFdata._fitData8(val,
data.wavelength,
data.data,
bounds=True,
tol=1e-8))
fig.canvas.draw_idle()
sfreq.on_changed(update)
plt.show()
def plotSIF5(shot, init=0, offset=-50):
#DATA GET
data = SIFdata.SIFData(shot)
fig, ax = plt.subplots()
plt.subplots_adjust(bottom=0.25)
#plt.axvline(ArHe_like_lambda['w'],color='b')
#plt.axvline(ArHe_like_lambda['x'],color='b')
#plt.axvline(ArHe_like_lambda['y'],color='b')
#plt.axvline(ArHe_like_lambda['z'],color='b')
#plt.axvline(ArLi_like_lambda['a'],color='r')
#plt.axvline(ArLi_like_lambda['j'],color='r')
#plt.axvline(ArLi_like_lambda['q'],color='r')
#plt.axvline(ArLi_like_lambda['r'],color='r')
#plt.axvline(ArLi_like_lambda['s'],color='r')
#plt.axvline(ArLi_like_lambda['t'],color='r')
plt.axvline(Wlambda['px375'], color='g')
temp = ['r', 'g', 'b', 'k', 'c', 'y']
inp = 0
temp2 = [['W', 43, 'c'], ['W', 44, 'c'], ['W', 45, 'c'], ['Ar', 16, 'b']]
for i in temp2:
print(i, temp[inp])
for j in SIFdata.lines[i[0]][i[1]]:
plt.axvline(j, color=i[2])
temp = scipy.sum(data.data, axis=0)
temp -= scipy.median(temp)
#argmax = scipy.argmax(temp[400:440])
#print(argmax)
#offset = -85650.6*Wlambda['px375'] - argmax + 34334.5 + 50 -400
#xdata = (laxis+offset/(-85650.6))
xdata = data.wavelength
ydata = temp / scipy.sum(temp)
inp1 = SIFdata.semiEmperical
#for i in inp1:
# plt.axvline(i,alpha=.4,color='k')
size = len(inp1)
if not bounds == None:
inp, bounds = SIFdata._assembleInit(xdata, ydata, bounds=bounds)
else:
inp = SIFdata._assembleInit(xdata, ydata)
testdata = SIFdata.gaussian(xdata, inp)
#plt.plot(xdata,testdata)
#plt.show()
#raise AttributeError
output = scipy.optimize.minimize(SIFdata.compare,
inp,
args=(xdata, ydata),
method=method,
bounds=bounds)
#l2, = plt.semilogy(laxis2+offset/(-85650.6), data.data[init], lw=2, color='green')
l, = plt.plot(data.wavelength,
sig.medfilt(data.data[init], 5),
lw=2,
color='red')
plt.ylim((0, 5e4)) #scipy.nanmax(data.data)))
axcolor = 'lightgoldenrodyellow'
axfreq = plt.axes([0.25, 0.1, 0.65, 0.03])
sfreq = DiscreteSlider(axfreq,
'Freq',
0,
len(data.time),
valinit=init,
valfmt='%0.0i',
increment=1.0)
def update(val):
freq = int(sfreq.val)
print(freq)
#l2.set_ydata(data.data[val])
l.set_ydata(sig.medfilt(data.data[val], 5))
fig.canvas.draw_idle()
sfreq.on_changed(update)
plt.show()
def plotSIF2(shot, init=0, offset=-50, waves=False, method=None, bounds=None):
#global ArHe_like_lambda
#DATA GET
data = SIFdata.SIFData(shot)
#plt.axvline(ArHe_like_lambda['w'],color='b')
#plt.axvline(ArHe_like_lambda['x'],color='b')1.7 #plt.axvline(ArHe_like_lambda['y'],color='b')
#plt.axvline(ArHe_like_lambda['z'],color='b')
#plt.axvline(ArLi_like_lambda['a'],color='r')
#plt.axvline(ArLi_like_lambda['j'],color='r')
#plt.axvline(ArLi_like_lambda['q'],color='r')
#plt.axvline(ArLi_like_lambda['r'],color='r')
#plt.axvline(ArLi_like_lambda['s'],color='r')
#plt.axvline(ArLi_like_lambda['t'],color='r')
#plt.axvline(Wlambda['px375'],color='g')
temp = ['r', 'g', 'b', 'k', 'c', 'y']
inp = 0
temp2 = [['W', 43, 'c'], ['W', 44, 'c'], ['W', 45, 'c'], ['Ar', 16, 'b']]
if waves:
plt.axvline(Wlambda['px375'], color='g')
for i in temp2:
print(i, temp[inp])
for j in SIFdata.lines[i[0]][i[1]]:
plt.axvline(j, color=i[2])
temp = scipy.sum(data.data, axis=0)
temp -= scipy.median(temp)
#argmax = scipy.argmax(temp[400:440])
#print(argmax)
#offset = -85650.6*Wlambda['px375'] - argmax + 34334.5 + 50 -400
#xdata = (laxis+offset/(-85650.6))
xdata = data.wavelength
ydata = temp / scipy.sum(temp)
inp1 = SIFdata.semiEmperical
#for i in inp1:
# plt.axvline(i,alpha=.4,color='k')
size = len(inp1)
#inp = scipy.zeros((3*size+1))
#inp[size:2*size] = inp1
#inp[2*size:3*size] = 1.2e-4*scipy.ones((len(inp1),))
#inp2 = inp[:-1].reshape((3,(len(inp)-1)/3))
#inp[-1] = -.0005
#inp[:size] = ydata[scipy.searchsorted(xdata, inp1)] - inp[-1]
#temp = (xdata - scipy.dot(scipy.ones((len(xdata),1)),scipy.atleast_2d(inp[1])).T).T/inp[2]
#inp[-1] = -.0005
#if not bounds == None:
# bounds = scipy.zeros((size*3+1,2))
# bounds[:size] = [0,1]
# bounds[size:2*size] = scipy.add(scipy.atleast_2d(inp[size:2*size]).T,[[-3e-5,3e-5]])
# print(scipy.add(scipy.atleast_2d(inp[size:2*size]).T,[[-5e-5,5e-5]]))
# bounds[2*size:3*size] = [0,3e-4]
# bounds[-1] = [-1,1]
#print(inp1)
#print(inp)
if not bounds == None:
inp, bounds = SIFdata._assembleInit(xdata, ydata, bounds=bounds)
else:
inp = SIFdata._assembleInit(xdata, ydata)
testdata = SIFdata.gaussian(xdata, inp)
#plt.plot(xdata,testdata)
#plt.show()
#raise AttributeError
output = scipy.optimize.minimize(SIFdata.compare2,
inp,
args=(xdata, ydata),
method=method,
bounds=bounds)
print(inp1 - output.x[:-1].reshape((3, size))[1])
print(output.x[:-1].reshape((3, size)))