def test_twiss():
xdb = Xdb(index_file='/home/iagapov/data/xdb/test/index.h5', mode='r')
f = xdb.read_undulator_config('sase3/1000eV')
exec(f)
lat = MagneticLattice(sase3)
#rematch(18.0, l_fodo, qdh, lat, extra_fodo, beam, qf, qd) # jeez...
tw0 = Twiss(beam)
print tw0
tws = twiss(lat, tw0, nPoints=1000)
f = plt.figure()
ax = f.add_subplot(111)
ax.set_xlim(0, lat.totalLen)
f.canvas.set_window_title('Betas [m]')
p1, = plt.plot(map(lambda p: p.s, tws),
map(lambda p: p.beta_x, tws),
lw=2.0)
p2, = plt.plot(map(lambda p: p.s, tws),
map(lambda p: p.beta_y, tws),
lw=2.0)
plt.grid(True)
plt.legend([p1, p2], [r'$\beta_x$', r'$\beta_y$', r'$D_x$'])
plt.show()
def test_print_parameters():
xdb = Xdb(index_file='/home/iagapov/data/xdb/test/index.h5', mode='r')
# available undulator configs
global g
g = xdb.file['Undulators']
for u in g.keys():
print u
for u2 in g[u].keys():
print ' ', u2
# available beams
global s
s = ''
def printname(name):
global s, g
if len(g[name].keys()) > 1:
s = s + '->' + name + '\n'
else:
s = s + '*' + name + '\n'
g = xdb.file['Beams']
g.visit(printname)
print s
def test_twiss():
xdb = Xdb(index_file='/home/iagapov/data/xdb/test/index.h5', mode='r')
f = xdb.read_undulator_config('sase3/1000eV')
exec(f)
lat = MagneticLattice(sase3)
#rematch(18.0, l_fodo, qdh, lat, extra_fodo, beam, qf, qd) # jeez...
tw0 = Twiss(beam)
print tw0
tws=twiss(lat, tw0, nPoints = 1000)
f=plt.figure()
ax = f.add_subplot(111)
ax.set_xlim(0, lat.totalLen)
f.canvas.set_window_title('Betas [m]')
p1, = plt.plot(map(lambda p: p.s, tws), map(lambda p: p.beta_x, tws), lw=2.0)
p2, = plt.plot(map(lambda p: p.s, tws), map(lambda p: p.beta_y, tws), lw=2.0)
plt.grid(True)
plt.legend([p1,p2], [r'$\beta_x$',r'$\beta_y$', r'$D_x$'])
plt.show()
def create_db(index_file):
#index_file = '/home/iagapov/data/xdb/test/index.h5'
print 'creating database', index_file
xdb = Xdb(index_file=index_file, mode='w')
xdb.create_index()
xdb.add_undulator(
'sase1', {
'input_file':
'/home/iagapov/workspace/xcode/repository/xfel/sase1/sase1.inp'
})
xdb.add_undulator_config(
'sase1', '24000eV/', {
'input_file':
'/home/iagapov/workspace/xcode/repository/xfel/sase1/sase1.inp'
})
xdb.add_undulator(
'sase3', {
'input_file':
'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'
})
xdb.add_undulator_config(
'sase3', '1000eV/', {
'input_file':
'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'
})
xdb.add_undulator_config(
'sase3', '1500eV/', {
'input_file':
'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'
})
xdb.add_undulator_config(
'sase3', '3000eV/', {
'input_file':
'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'
})
xdb.add_beam('17.5GeV/nC/80fs/', {})
xdb.file.close()
print 'done'
def extract(path, ran, h5_file=None):
h = 4.135667516e-15
c = 299792458.0
verbose = False
powers_z = []
spectra = []
pulses = []
total_power = []
slice_files = []
for run_id in ran:
run_dir = path + '/run_' + str(run_id)
t1 = time.time()
output_file = run_dir + '/run.' + str(run_id) + '.gout'
if rank == 0:
print 'extracting run ', run_id
g = readGenesisOutput(output_file)
else:
g = None
g = comm.bcast(g, root=0)
t2 = time.time()
if verbose: print 'file reading:', t2 - t1, ' sec'
npoints = g('ncar')
zstop = g('zstop')
delz = g('delz')
xlamd = g('xlamd')
xlamds = g('xlamds')
nslice = int(g('nslice'))
dgrid = g('dgrid')
zsep = g('zsep')
E_gamma = h * c / xlamds
power_z = get_power_z(g)
pulse, t = get_power_exit(g)
powers_z.append(power_z)
#slices = readRadiationFile(fileName=output_file + '.dfl', npoints=npoints)
slices = readRadiationFile_mpi(comm=comm,
fileName=output_file + '.dfl',
npoints=npoints)
slice_files.append(output_file + '.dfl')
if rank == 0:
P = np.zeros_like(slices[:, 0, 0])
E = np.zeros_like(slices[:, 0, 0])
for i in xrange(len(P)):
P[i] = sum(
np.multiply(np.abs(slices[i, :, :]),
np.abs(slices[i, :, :])))
#s = sum( np.multiply(slices[i,:,:]slices[i,:,:])
#P[i] = abs(s*s.conjugate()) #* (1 / npoints)**2
E[i] = sum(slices[i, :, :])
pulses.append(P)
spec = fft(E)
spec = np.sqrt(spec * np.conj(spec))
spec = np.real(np.roll(spec, len(spec) / 2))
tmax = nslice * zsep * xlamds / c
freq_ev = fftfreq(len(spec), d=zsep * xlamds / c)
freq_ev = h * np.roll(freq_ev, len(spec) / 2)
spectra.append(np.real(spec))
total_power.append(sum(P) * zsep * xlamds / c)
del slices
if rank == 0:
t = 1.0e+15 * zsep * xlamds * np.arange(0, len(pulses[0])) / c
fig = figure()
fig.add_subplot(111)
pz_mean, pz_std, pz_med, pz_worst, imed, iworst = stats(powers_z)
plot(g.z, pz_med, 'b', lw=3)
errorbar(g.z, pz_mean, yerr=pz_std, fmt='r--', lw=1, capsize=7)
fig = figure()
fig.add_subplot(111)
for i in xrange(len(pulses)):
plot(t, pulses[i], 'b--', lw=1)
p_mean, p_std, p_med, p_worst, imed, iworst = stats(pulses)
slices = readRadiationFile(fileName=slice_files[imed], npoints=npoints)
slices_2d = np.zeros(slices[0, :, :].shape, dtype=complex)
#print slices_2d.shape
for i in xrange(slices.shape[0]):
slices_2d[:, :] += slices[i, :, :]
plot(t, p_med, 'g', lw=3)
errorbar(t, p_mean, yerr=p_std, fmt='r--', lw=1, capsize=7)
n_gamma = np.sum(p_mean)
mu1, mu2, sig1, sig2, sig12 = fit_gauss_2d(
np.linspace(-dgrid / 2.0, dgrid / 2.0, npoints),
np.linspace(-dgrid / 2.0, dgrid / 2.0, npoints), slices_2d)
field_sig_x, field_sig_y = sig1, sig2
fig = figure()
fig.add_subplot(111)
for i in xrange(len(spectra)):
plot(freq_ev, spectra[i], 'b--', lw=1)
s_mean, s_std, s_med, s_worst, imed, iworst = stats(spectra)
plot(freq_ev, s_med, 'g', lw=3)
#errorbar(t, p_mean, yerr=p_std, fmt='r--',lw=1, capsize=7)
figure()
n, bins = histogram(total_power / mean(total_power), 40, normed=False)
bar(bins[1:] - (bins[1] - bins[0]) / 2.0,
n,
width=(bins[1] - bins[0]),
alpha=0.5)
if h5_file != None:
print 'writing to ', h5_file
xdb = Xdb(index_file=h5_file, mode='w')
params = {}
params['n_gamma'] = n_gamma
params['power_z_mean'] = pz_mean
params['power_z_med'] = pz_med
params['power_z_std'] = pz_std
params['pulse_mean'] = np.real(p_mean)
params['pulse_med'] = np.real(p_med)
params['pulse_std'] = np.real(p_std)
params['spec_mean'] = np.real(s_mean)
params['spec_med'] = np.real(s_med)
params['spec_std'] = np.real(s_std)
params['total_power'] = np.real(total_power)
params['field_med'] = slices_2d
params['field_sig_x'] = field_sig_x
params['field_sig_y'] = field_sig_y
params['t'] = np.real(t)
params['z'] = np.array(g.z)
params['I'] = np.array(g.I)
params['freq_ev'] = np.real(freq_ev)
#params['version'] = 'v0'
params['dgrid'] = dgrid
params['e_gamma'] = E_gamma
xdb.add_fel_calculation('sase/', params, root='/')
show()
def plot_stats(idx='', base='', root='/FEL/'):
xdb = Xdb(index_file=idx, mode='r')
try:
fig = figure(idx)
except:
fig = figure()
ax = fig.add_subplot(111)
ax.grid(True)
ax.set_xlabel('[fs]')
ax.set_ylabel('[W]')
p_med = np.array(xdb.file[root + base + '/pulse_med'])
p_mean = np.array(xdb.file[root + base + '/pulse_mean'])
p_std = np.array(xdb.file[root + base + '/pulse_std'])
t = np.array(xdb.file[root + base + '/t'])
I = xdb.file[root + base + '/I']
I = np.concatenate([I, np.zeros(len(t) - len(I))])
p1, = ax.plot(t, p_med, 'g-', lw=2)
p2, = ax.plot(t, p_mean, 'b-', lw=5)
fill_between(t, p_mean - p_std, p_mean + p_std, alpha=0.2)
#p2, __ , __ , = ax.errorbar(t[::10], p_mean[::10], yerr=p_std[::10], fmt='g--',lw=3, capsize=5)
# gauss fit
mu, sig = fit_gauss_1d(t, p_mean)
sig2 = fwhm(t, p_mean)
print 'pulse duration, fs, (2*rms/ FWHM): ', 2 * sig, sig2
E_pulse = np.sum(p_med) * (1.e-15 * (t[-1] - t[0])) / len(t)
print 'Pulse energy [mJ]: ', E_pulse * 1.e3
print 'Peak power (mean) [GW]: ', np.max(p_mean) / 1.e9
print 'Peak power (med) [GW]: ', np.max(p_med) / 1.e9
try:
E_gamma = xdb.file[root + base].attrs["e_gamma"]
Ng = E_pulse / 1.6e-19 / E_gamma
print 'N photons: ', Ng / 1.e12, ' x 10^12'
except:
E_gamma = 1.e24
print 'N photons: photon energy undefined!!!'
sig = sig2 / 2
#p3, = ax.plot(t, np.max(p_mean) * np.exp(-(t-mu)**2 / (2.*sig**2)), 'r--', lw=4)
#plot(t, np.max(p_med) * (I / np.max(I)), 'b--', lw=3)
'''
figure()
plot(t, I, 'b--', lw=3)
print 'integral', np.sum(I) * (t[1]-t[0])
'''
#p3, = ax.plot(t, np.sum(p_mean) * (t[-1] - t[0]) / len(t) / (np.sqrt(2*pi) * sig) * np.exp(-(t-mu)**2 / (2*sig**2)), 'r--', lw=3)
#ax.legend([p1,p2,p3], ['median','mean/std','gauss fit'])
fig = figure()
ax = fig.add_subplot(111)
ax.grid(True)
ax.set_xlabel('[eV]')
ax.set_ylabel('[A.U.]')
s_med = np.array(xdb.file[root + base + '/spec_med'])
s_mean = np.array(xdb.file[root + base + '/spec_mean'])
s_std = np.array(xdb.file[root + base + '/spec_std'])
t = np.array(xdb.file[root + base + '/freq_ev'])
ax.plot(t, s_med, 'g-', lw=2)
ax.plot(t, s_mean, 'b-', lw=5)
mu, sig = fit_gauss_1d(t, p_mean)
#p2, = ax.plot(t, p_mean, '-', lw=3)
#p3, = ax.plot(t, np.max(s_mean) * np.exp(-(t-mu)**2 / (2.*sig**2)), 'r--', lw=4)
#ax.errorbar(t[::10], s_mean[::10], yerr=s_std[::10], fmt='g--',lw=2, capsize=5, alpha=0.9)
fill_between(t, s_mean - s_std, s_mean + s_std, alpha=0.2)
# gauss fit
mu, sig = fit_gauss_1d(t, s_mean)
sig2 = fwhm(t, s_mean)
print 'pulse bandwidth, eV, (2*rms/ FWHM): ', 2 * sig, sig2, '(', 2 * 100 * sig / E_gamma, 100 * sig2 / E_gamma, '%)'
print 'Max brightness (mean) [AU]: ', np.max(s_mean) / 1.e8
print 'Max brightness (med) [AU]: ', np.max(s_med) / 1.e8
X = E_gamma * 1.e-3
print '0.1%bw=', X
I = np.sum(s_mean) * (t[1] - t[0])
print 'I [x 10^12] =', I / 1.e12
Y = Ng / I
print 'Y', Ng / (I)
print 'Max brightness (mean) [N_phot / 0.1% bw]: ', np.max(s_mean) * Y * X
sig = sig2 / 2
#ax.plot(t, np.max(s_mean) * np.exp(-(t-mu)**2 / (2.*sig**2)), 'r--', lw=3)
#ax.plot(t, np.max(s_mean) * np.exp(-(t-mu)**2 / (2.*sig2**2)), 'r--', lw=3)
#ax.plot(t, np.sum(s_mean) * (t[-1] - t[0]) / len(t) / (np.sqrt(2.*pi) * sig) * np.exp(-(t-mu)**2 / (2.*sig**2)), 'r--', lw=3)
#fig = figure()
#ax = fig.add_subplot(323)
fig = figure()
ax = fig.add_subplot(111)
f_med = xdb.file[root + base + '/field_med']
try:
dgrid = xdb.file[root + base].attrs['dgrid']
except:
print 'dgrid not defined'
dgrid = 1
ax.imshow(np.abs(f_med),
extent=[
-1.e3 * dgrid / 2, 1.e3 * dgrid / 2, -1.e3 * dgrid / 2,
1.e3 * dgrid / 2
],
aspect='auto',
cmap='YlOrRd')
ax.set_ylabel('[mm]')
ax.set_xlabel('[mm]')
#ax = fig.add_subplot(324)
fig = figure()
ax = fig.add_subplot(111)
ax.set_ylabel('A.U.')
ax.set_xlabel('[mm]')
ax.set_yticks([])
P = np.abs(f_med[:, int(f_med.shape[0] / 2)])
x = np.linspace(-1.e3 * dgrid / 2.0, 1.e3 * dgrid / 2.0, len(P))
ax.plot(x, P, lw=3)
# gauss fit
mu, sig = fit_gauss_1d(x, P)
sig2 = fwhm(x, P)
#sig /= 1.8
print 'spot size, mu m, (2*rms / FWHM): ', 2 * sig / 1.e-6, '/', sig2 / 1.e-6
#p3, = ax.plot(x, np.sum(P) * (x[-1] - x[0]) / len(x) / (np.sqrt(2*pi) * sig) * np.exp(-(x-mu)**2 / (2*sig**2)), 'r--', lw=3)
ax.grid(True)
#ax = fig.add_subplot(325)
fig = figure()
ax = fig.add_subplot(111)
ax.set_ylabel('A.U.')
ax.set_xlabel('[mrad]')
ax.set_yticks([])
P = np.abs(f_med[:, int(f_med.shape[0] / 2)])
x = np.linspace(-1.e3 * dgrid / 2.0, 1.e3 * dgrid / 2.0, len(P))
ax.plot(x, P, lw=3)
# gauss fit
mu, sig = fit_gauss_1d(x, P)
sig2 = fwhm(x, P)
#sig /= 1.8
print 'divergence, mu rad, (2*rms / FWHM): ', 2 * sig / 1.e-6, '/', sig2 / 1.e-6
#p3, = ax.plot(x, np.sum(P) * (x[-1] - x[0]) / len(x) / (np.sqrt(2*pi) * sig) * np.exp(-(x-mu)**2 / (2*sig**2)), 'r--', lw=3)
#ax = fig.add_subplot(326)
fig = figure()
ax = fig.add_subplot(111)
ax.grid(True)
ax.set_ylabel('Pulse energy [mJ]')
ax.set_xlabel('[m]')
#ax.set_yticks([])
z = np.array(xdb.file[root + base + '/z'])
power_z = np.array(xdb.file[root + base + '/power_z_mean'])
power_z_std = np.array(xdb.file[root + base + '/power_z_std'])
fact = E_pulse * 1.e3 / power_z[-1]
power_z *= fact
power_z_std *= fact
ax.plot(z, power_z, lw=4)
fill_between(z, power_z - power_z_std, power_z + power_z_std, alpha=0.3)
#p2, __ , __ , = ax.errorbar(z, power_z, yerr=power_z_std, fmt='g--',lw=3, capsize=5)
'''
fig = figure(base + 'HISTO')
total_power = xdb.file[root + base + '/total_power']
n, bins = histogram( (total_power/ mean(total_power) - 1) * 10 + 1, 50, normed=False)
bar( bins[1:] - (bins[1]-bins[0])/2.0 , n / 4.0, width = (bins[1]-bins[0]), alpha=0.5)
'''
def distr(M, E):
t = np.linspace(E / 20, E * 5, 1000)
return t, pow(M, M) / gamma(M) * pow(t / E, M - 1) / E * exp(
-M * t / E)
#t, E = distr(10.1, 1)
#plot(t, E, '--', lw=3), grid(True)
show()
def create_db(index_file):
#index_file = '/home/iagapov/data/xdb/test/index.h5'
print 'creating database', index_file
xdb = Xdb(index_file=index_file, mode='w')
xdb.create_index()
xdb.add_undulator('sase1', {'input_file':'/home/iagapov/workspace/xcode/repository/xfel/sase1/sase1.inp'})
xdb.add_undulator_config('sase1', '24000eV/', {'input_file':'/home/iagapov/workspace/xcode/repository/xfel/sase1/sase1.inp'})
xdb.add_undulator('sase3', {'input_file':'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'})
xdb.add_undulator_config('sase3', '1000eV/', {'input_file':'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'})
xdb.add_undulator_config('sase3', '1500eV/', {'input_file':'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'})
xdb.add_undulator_config('sase3', '3000eV/', {'input_file':'/home/iagapov/workspace/xcode/repository/xfel/sase3/sase3.inp'})
xdb.add_beam('17.5GeV/nC/80fs/',{})
xdb.file.close()
print 'done'
def extract(path, ran, h5_file = None):
h = 4.135667516e-15
c = 299792458.0
verbose = False
powers_z = []
spectra = []
pulses = []
total_power = []
slice_files = []
for run_id in ran:
run_dir = path + '/run_' + str(run_id)
t1 = time.time()
output_file = run_dir + '/run.' + str(run_id) + '.gout'
if rank == 0:
print 'extracting run ', run_id
g = readGenesisOutput(output_file)
else:
g = None
g = comm.bcast(g, root=0)
t2 = time.time()
if verbose: print 'file reading:', t2-t1, ' sec'
npoints = g('ncar')
zstop = g('zstop')
delz = g('delz')
xlamd = g('xlamd')
xlamds = g('xlamds')
nslice = int(g('nslice'))
dgrid = g('dgrid')
zsep = g('zsep')
E_gamma = h * c / xlamds
power_z = get_power_z(g)
pulse, t = get_power_exit(g)
powers_z.append(power_z)
#slices = readRadiationFile(fileName=output_file + '.dfl', npoints=npoints)
slices = readRadiationFile_mpi(comm=comm, fileName=output_file+'.dfl', npoints=npoints)
slice_files.append(output_file+'.dfl')
if rank == 0:
P = np.zeros_like(slices[:,0,0])
E = np.zeros_like(slices[:,0,0])
for i in xrange(len(P)):
P[i] = sum( np.multiply( np.abs(slices[i,:,:]), np.abs(slices[i,:,:]) ) )
#s = sum( np.multiply(slices[i,:,:]slices[i,:,:])
#P[i] = abs(s*s.conjugate()) #* (1 / npoints)**2
E[i] = sum( slices[i,:,:])
pulses.append(P)
spec = fft(E)
spec = np.sqrt( spec * np.conj(spec) )
spec = np.real( np.roll(spec, len(spec)/2) )
tmax = nslice * zsep * xlamds / c
freq_ev = fftfreq(len(spec), d=zsep * xlamds / c)
freq_ev = h*np.roll(freq_ev, len(spec)/2)
spectra.append( np.real(spec) )
total_power.append( sum(P) * zsep * xlamds / c )
del slices
if rank == 0:
t = 1.0e+15 * zsep * xlamds * np.arange(0,len(pulses[0])) / c
fig = figure()
fig.add_subplot(111)
pz_mean, pz_std, pz_med, pz_worst, imed, iworst = stats(powers_z)
plot(g.z, pz_med, 'b', lw=3)
errorbar(g.z, pz_mean, yerr=pz_std, fmt='r--',lw=1, capsize=7)
fig = figure()
fig.add_subplot(111)
for i in xrange(len(pulses)):
plot(t, pulses[i], 'b--', lw=1)
p_mean, p_std, p_med, p_worst, imed, iworst = stats(pulses)
slices = readRadiationFile(fileName=slice_files[imed], npoints=npoints)
slices_2d = np.zeros(slices[0,:,:].shape, dtype=complex)
#print slices_2d.shape
for i in xrange(slices.shape[0]):
slices_2d[:,:] += slices[i,:,:]
plot(t, p_med, 'g', lw=3)
errorbar(t, p_mean, yerr=p_std, fmt='r--',lw=1, capsize=7)
n_gamma = np.sum(p_mean)
mu1,mu2, sig1, sig2, sig12 = fit_gauss_2d(np.linspace(-dgrid/2.0,dgrid/2.0, npoints),np.linspace(-dgrid/2.0,dgrid/2.0, npoints), slices_2d)
field_sig_x, field_sig_y = sig1, sig2
fig = figure()
fig.add_subplot(111)
for i in xrange(len(spectra)):
plot(freq_ev, spectra[i], 'b--', lw=1)
s_mean, s_std, s_med, s_worst, imed, iworst = stats(spectra)
plot(freq_ev, s_med, 'g', lw=3)
#errorbar(t, p_mean, yerr=p_std, fmt='r--',lw=1, capsize=7)
figure()
n, bins = histogram(total_power / mean(total_power), 40, normed=False)
bar( bins[1:] - (bins[1]-bins[0])/2.0 , n, width = (bins[1]-bins[0]), alpha=0.5)
if h5_file != None:
print 'writing to ', h5_file
xdb = Xdb(index_file=h5_file, mode='w')
params = {}
params['n_gamma'] = n_gamma
params['power_z_mean'] = pz_mean
params['power_z_med'] = pz_med
params['power_z_std'] = pz_std
params['pulse_mean'] = np.real(p_mean)
params['pulse_med'] = np.real(p_med)
params['pulse_std'] = np.real(p_std)
params['spec_mean'] = np.real(s_mean)
params['spec_med'] = np.real(s_med)
params['spec_std'] = np.real(s_std)
params['total_power'] = np.real(total_power)
params['field_med'] = slices_2d
params['field_sig_x'] = field_sig_x
params['field_sig_y'] = field_sig_y
params['t'] = np.real(t)
params['z'] = np.array(g.z)
params['I'] = np.array(g.I)
params['freq_ev'] = np.real(freq_ev)
#params['version'] = 'v0'
params['dgrid'] = dgrid
params['e_gamma'] = E_gamma
xdb.add_fel_calculation('sase/', params, root='/')
show()