Skip to content

Commit

Permalink
Merge branch 'develop' into feature_cross_measurement_dephasing_measu…
Browse files Browse the repository at this point in the history 
…rements_from_device_object
  • Loading branch information
@AdriaanRol
AdriaanRol committed May 14, 2018
2 parents 47630c2 + e2f9497 commit 24a0611
Show file tree
Hide file tree
Showing 34 changed files with 286 additions and 105 deletions.
15 changes: 11 additions & 4 deletions pycqed/analysis/fitting_models.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -802,8 +802,15 @@ def gauss_2D_guess(model, data, x, y):
y_proj_data = np.mean(data_grid, axis=1)
x.sort()
y.sort()
x_guess = lmfit.models.GaussianModel().guess(data=x_proj_data, x=np.unique(x))
y_guess = lmfit.models.GaussianModel().guess(data=y_proj_data, x=np.unique(y))
xm = lmfit.models.GaussianModel()
ym = lmfit.models.GaussianModel()
x_guess = xm.guess(data=x_proj_data, x=np.unique(x))
x_res = xm.fit(data=x_proj_data, x=np.unique(x), params=x_guess)
y_guess = ym.guess(data=y_proj_data, x=np.unique(y))
y_res = ym.fit(data=y_proj_data, x=np.unique(y), params=y_guess)

x_guess = x_res.params
y_guess = y_res.params
model.set_param_hint('amplitude', value=amp, min=0.9*amp, max=1.1*amp,
vary=True)
model.set_param_hint('sigma_x', value=x_guess['sigma'].value, min=0,
Expand Down Expand Up @@ -924,8 +931,8 @@ def ro_double_gauss_guess(model, data, x, fixed_p01 = False, fixed_p10 = False):
amp1 = 0.99 * np.max(data[1]) * guess1['sigma'] * f
spurious0 = max(1-(amp0/intarea0), 1e-3)
spurious1 = max(1-(amp1/intarea1), 1e-3)
p01 = fixed_p01 or spurious0
p10 = fixed_p10 or spurious1
p01 = fixed_p01 if fixed_p01 is not False else spurious0
p10 = fixed_p10 if fixed_p10 is not False else spurious1
model.set_param_hint('A_spurious', value=p01, min=0, max=1,
vary=fixed_p01 is False)
model.set_param_hint('B_spurious', value=p10, min=0, max=1,
Expand Down
8 changes: 7 additions & 1 deletion pycqed/analysis/measurement_analysis.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2987,8 +2987,14 @@ class SSRO_Analysis(MeasurementAnalysis):
the optimum threshold and fidelity from cumulative histograms.
'''

def __init__(self, rotate=True, close_fig=True, channels=['I', 'Q'],
def __init__(self, rotate=True, close_fig=True, channels=None,
hist_log_scale: bool = True, **kw):
if channels is None:
channels = ['I', 'Q']

logging.warning('The use of this class is deprectated!' +
' Use the new v2 analysis instead.')

kw['h5mode'] = 'r+'
self.rotate = rotate
self.channels = channels
Expand Down
2 changes: 2 additions & 0 deletions pycqed/analysis_v2/quantum_efficiency_analysis.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -641,6 +641,7 @@ def __init__(self, t_start: str = None, t_stop: str = None, label: str = '_depha
def extract_data(self):
# Load data
super().extract_data()
#todo: we need an option to remove outliers and the reference point

# Set output paths
youngest = np.max(self.raw_data_dict['datetime'])
Expand Down Expand Up @@ -846,6 +847,7 @@ def extract_data(self):
# Load data
super().extract_data()

#todo: we need an option to remove outliers and the reference point
# Set output paths
youngest = np.max(self.raw_data_dict['datetime'])
youngest += datetime.timedelta(seconds=1)
Expand Down
176 changes: 91 additions & 85 deletions pycqed/analysis_v2/readout_analysis.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -28,10 +28,9 @@
class Singleshot_Readout_Analysis(ba.BaseDataAnalysis):

def __init__(self, t_start: str=None, t_stop: str=None,
label: str='',
label: str='', do_fitting: bool = True,
data_file_path: str=None,
options_dict: dict=None, extract_only: bool=False,
do_fitting: bool=True, auto=True):
options_dict: dict=None, auto=True, **kw):
'''
options dict options:
'fixed_p10' fixes p(e|g) (do not vary in fit)
Expand All @@ -51,10 +50,10 @@ def __init__(self, t_start: str=None, t_stop: str=None,
see BaseDataAnalysis for more.
'''
super().__init__(t_start=t_start, t_stop=t_stop,
label=label,
label=label, do_fitting=do_fitting,
data_file_path=data_file_path,
options_dict=options_dict,
extract_only=extract_only, do_fitting=do_fitting)
**kw)
self.single_timestamp = True
self.params_dict = {
'measurementstring': 'measurementstring',
Expand All @@ -68,7 +67,6 @@ def __init__(self, t_start: str=None, t_stop: str=None,
man_angle = self.options_dict.get('rotation_angle', False) is False
self.options_dict['auto_rotation_angle'] = self.options_dict.get('auto_rotation_angle', man_angle)


if auto:
self.run_analysis()

Expand Down Expand Up @@ -111,13 +109,14 @@ def process_data(self):
data_range_y = (np.min([np.min(b) for b in shots[:, 1]]),
np.max([np.max(b) for b in shots[:, 1]]))
data_range_xy = (data_range_x, data_range_y)
nr_bins_2D = self.options_dict.get('nr_bins_2D', 6*np.sqrt(nr_bins))
H0, xedges, yedges = np.histogram2d(x=shots[0, 0],
y=shots[0, 1],
bins=4*np.sqrt(nr_bins),
bins=nr_bins_2D,
range=data_range_xy)
H1, xedges, yedges = np.histogram2d(x=shots[1, 0],
y=shots[1, 1],
bins=4*np.sqrt(nr_bins),
bins=nr_bins_2D,
range=data_range_xy)
binsize_x = xedges[1] - xedges[0]
binsize_y = yedges[1] - yedges[0]
Expand All @@ -138,35 +137,30 @@ def process_data(self):
independent_vars=['x', 'y'])
gauss2D_model_1 = lmfit.Model(gaussian_2D,
independent_vars=['x', 'y'])
guess0 = gauss_2D_guess(model=gauss2D_model_0,
data=H0,
x=bin_centers_x,
y=bin_centers_y)
guess1 = gauss_2D_guess(model=gauss2D_model_1,
data=H1,
x=bin_centers_x,
y=bin_centers_y)

x2d=np.array([bin_centers_x]*len(bin_centers_y))
y2d=np.array([bin_centers_y]*len(bin_centers_x)).transpose()
fitres0 = gauss2D_model_0.fit(data=H0,
x=x2d,
y=y2d,
guess0 = gauss_2D_guess(model=gauss2D_model_0, data=H0.transpose(),
x=bin_centers_x, y=bin_centers_y)
guess1 = gauss_2D_guess(model=gauss2D_model_1, data=H1.transpose(),
x=bin_centers_x, y=bin_centers_y)

x2d = np.array([bin_centers_x]*len(bin_centers_y))
y2d = np.array([bin_centers_y]*len(bin_centers_x)).transpose()
fitres0 = gauss2D_model_0.fit(data=H0.transpose(), x=x2d, y=y2d,
**guess0)
fitres1 = gauss2D_model_1.fit(data=H1,
x=x2d,
y=y2d,
fitres1 = gauss2D_model_1.fit(data=H1.transpose(), x=x2d, y=y2d,
**guess1)

fr0 = fitres0.best_values
fr1 = fitres1.best_values
x0 = fr0['center_x']
x1 = fr1['center_x']
y0 = fr0['center_y']
y1 = fr1['center_y']

x0 = fitres0.best_values['center_x']
x1 = fitres1.best_values['center_x']
y0 = fitres0.best_values['center_y']
y1 = fitres1.best_values['center_y']
self.proc_data_dict['IQ_pos'] = [[x0, x1], [y0, y1]]
dx = x1 - x0
dy = y1 - y0
mid = [x0 + dx/2, y0 + dy/2]
angle = -np.arctan2(dy, dx)-np.pi/2
angle = np.arctan2(dy, dx)
else:
mid = [0, 0]

Expand All @@ -178,8 +172,8 @@ def process_data(self):

self.proc_data_dict['raw_offset'] = [*mid, angle]
# create matrix
rot_mat = [[+np.cos(angle), -np.sin(angle)],
[+np.sin(angle), +np.cos(angle)]]
rot_mat = [[+np.cos(-angle), -np.sin(-angle)],
[+np.sin(-angle), +np.cos(-angle)]]
# rotate data accordingly
eff_sh = np.zeros(len(shots[0]), dtype=np.ndarray)
eff_sh[0] = np.dot(rot_mat[0], shots[0])# - mid
Expand Down Expand Up @@ -238,7 +232,8 @@ def process_data(self):
# Average assignment fidelity: F_ass = (P01 - P10 )/2
# where Pxy equals probability to measure x when starting in y
F_vs_th = (1-(1-abs(necumsum0 - necumsum1))/2)
opt_idx = np.argmax(F_vs_th)
opt_idxs = np.argwhere(F_vs_th == np.amax(F_vs_th))
opt_idx = int(round(np.average(opt_idxs)))
self.proc_data_dict['F_assignment_raw'] = F_vs_th[opt_idx]
self.proc_data_dict['threshold_raw'] = all_x[opt_idx]

Expand Down Expand Up @@ -308,7 +303,7 @@ def infid_vs_th(x):
self._CDF_1 = CDF_1
self._infid_vs_th = infid_vs_th

thr_guess = (bv['B_center'] - bv['A_center'])/2
thr_guess = (3*bv['B_center'] - bv['A_center'])/2
opt_fid = minimize(infid_vs_th, thr_guess)

# for some reason the fit sometimes returns a list of values
Expand All @@ -327,17 +322,17 @@ def infid_vs_th(x):

def CDF_0_discr(x):
return gaussianCDF(x, amplitude=1,
mu=bv['A_center'], sigma=bv['A_sigma'])
mu=bv['A_center'], sigma=bv['A_sigma'])

def CDF_1_discr(x):
return gaussianCDF(x, amplitude=1,
mu=bv['B_center'], sigma=bv['B_sigma'])
mu=bv['B_center'], sigma=bv['B_sigma'])

def disc_infid_vs_th(x):
cdf0 = gaussianCDF(x, amplitude=1, mu=bv['A_center'],
sigma=bv['A_sigma'])
sigma=bv['A_sigma'])
cdf1 = gaussianCDF(x, amplitude=1, mu=bv['B_center'],
sigma=bv['B_sigma'])
sigma=bv['B_sigma'])
return (1-np.abs(cdf0 - cdf1))/2

self._CDF_0_discr = CDF_0_discr
Expand All @@ -354,6 +349,11 @@ def disc_infid_vs_th(x):

self.proc_data_dict['threshold_discr'] = opt_fid_discr['x'][0]

fr = self.fit_res['shots_all']
bv = fr.params
self.proc_data_dict['residual_excitation'] = bv['B_spurious'].value
self.proc_data_dict['measurement_induced_relaxation'] = bv['A_spurious'].value

def prepare_plots(self):
# Did we load two voltage components (shall we do 2D plots?)
two_dim_data = len(self.proc_data_dict['all_channel_int_voltages'][0]) == 2
Expand Down Expand Up @@ -414,6 +414,7 @@ def prepare_plots(self):
cdf_ys = self.proc_data_dict['cumsum_y']
cdf_ys[0] = cdf_ys[0]/np.max(cdf_ys[0])
cdf_ys[1] = cdf_ys[1]/np.max(cdf_ys[1])
xra = (bin_x[0], bin_x[-1])

self.plot_dicts['cdf_shots_0'] = {
'title': 'Culmulative Shot Counts (no binning)' + title,
Expand All @@ -422,6 +423,7 @@ def prepare_plots(self):
'xvals': cdf_xs[0],
'yvals': cdf_ys[0],
'setlabel': label_0,
'xrange': xra,
'line_kws': {'color': 'C0', 'alpha': 0.3},
'marker': '',
'xlabel': eff_voltage_label,
Expand Down Expand Up @@ -470,8 +472,8 @@ def prepare_plots(self):
iq_centers = self.proc_data_dict['IQ_pos']
peak_marker_2D = {
'plotfn': self.plot_line,
'xvals': iq_centers[0],
'yvals': iq_centers[1],
'xvals': iq_centers[1],
'yvals': iq_centers[0],
'xlabel': x_volt_label,
'xunit': x_volt_unit,
'ylabel': y_volt_label,
Expand All @@ -483,21 +485,26 @@ def prepare_plots(self):
'setlabel': 'Peaks',
'do_legend': True,
}
self.plot_dicts['2D_histogram_0'] = {
'title': 'Raw '+label_0+' Binned Shot Counts' + title,
'ax_id': '2D_histogram_0',
'plotfn': self.plot_colorxy,
'xvals': self.proc_data_dict['2D_histogram_y'],
'yvals': self.proc_data_dict['2D_histogram_x'],
'zvals': self.proc_data_dict['2D_histogram_z'][0],
'xlabel': x_volt_label,
'xunit': x_volt_unit,
'ylabel': y_volt_label,
'yunit': y_volt_unit,
'zlabel': z_hist_label,
'zunit': '-',
'cmap': 'Blues',
}
peak_marker_2D_rot = deepcopy(peak_marker_2D)
peak_marker_2D_rot['xvals'] = iq_centers[0]
peak_marker_2D_rot['yvals'] = iq_centers[1]

self.plot_dicts['2D_histogram_0'] = {
'title': 'Raw '+label_0+' Binned Shot Counts' + title,
'ax_id': '2D_histogram_0',
'plotfn': self.plot_colorxy,
'xvals': self.proc_data_dict['2D_histogram_y'],
'yvals': self.proc_data_dict['2D_histogram_x'],
'zvals': self.proc_data_dict['2D_histogram_z'][0],
'xlabel': x_volt_label,
'xunit': x_volt_unit,
'ylabel': y_volt_label,
'yunit': y_volt_unit,
'zlabel': z_hist_label,
'zunit': '-',
'cmap': 'Blues',
}
if iq_centers is not None:
dp = deepcopy(peak_marker_2D)
dp['ax_id'] = '2D_histogram_0'
self.plot_dicts['2D_histogram_0_marker'] = dp
Expand All @@ -524,19 +531,19 @@ def prepare_plots(self):

#### Scatter Shots
volts = self.proc_data_dict['all_channel_int_voltages']
vxr = [np.min([np.min(a) for a in volts[:][0]]),
np.max([np.max(a) for a in volts[:][0]])]
vyr = [np.min([np.min(a) for a in volts[:][1]]),
vxr = [np.min([np.min(a) for a in volts[:][1]]),
np.max([np.max(a) for a in volts[:][1]])]
vyr = [np.min([np.min(a) for a in volts[:][0]]),
np.max([np.max(a) for a in volts[:][0]])]
self.plot_dicts['2D_shots_0'] = {
'title': 'Raw Shots' + title,
'ax_id': '2D_shots',
'plotfn': self.plot_line,
'xvals': volts[0][1],
'yvals': volts[0][0],
'range': [vxr, vyr],
'xrange': vxr,
'yrange': vyr,
#'range': [vxr, vyr],
#'xrange': vxr,
#'yrange': vyr,
'xlabel': x_volt_label,
'xunit': x_volt_unit,
'ylabel': y_volt_label,
Expand All @@ -554,9 +561,9 @@ def prepare_plots(self):
'plotfn': self.plot_line,
'xvals': volts[1][1],
'yvals': volts[1][0],
'range': [vxr, vyr],
'xrange': vxr,
'yrange': vyr,
#'range': [vxr, vyr],
#'xrange': vxr,
#'yrange': vyr,
'xlabel': x_volt_label,
'xunit': x_volt_unit,
'ylabel': y_volt_label,
Expand Down Expand Up @@ -632,7 +639,6 @@ def prepare_plots(self):
'do_legend': True,
}


##########################################
# Add textbox (eg.g Thresholds, fidelity #
# information, number of shots etc) #
Expand Down Expand Up @@ -671,31 +677,31 @@ def prepare_plots(self):
'{:.1f}%'.format(
self.proc_data_dict['F_discr']*100))
fit_text += fit_th_msg
snr = self.fit_res['shots_all'].params['SNR']
fit_text += '\nSNR (fit) = ${:.3f}\\pm{:.3f}$'.format(snr.value, snr.stderr)

fr = self.fit_res['shots_all']
bv = fr.params
a_sp = bv['A_spurious']
fit_text += '\n\nSpurious Excitations:'
fit_text += '\n$p(e|0) = {:.3f}$'.format(a_sp.value)
if self.options_dict.get('fixed_p01', True) == True:
fit_text += '$\\pm{:.3f}$'.format(a_sp.stderr)
else:
fit_text += ' (fixed)'

b_sp = bv['B_spurious']
fit_text += ' \n$p(g|\\pi) = {:.3f}$'.format(b_sp.value)
if self.options_dict.get('fixed_p10', True) == True:
fit_text += '$\\pm{:.3f}$'.format(b_sp.stderr)
else:
fit_text += ' (fixed)'
snr = self.fit_res['shots_all'].params['SNR']
fit_text += '\nSNR (fit) = ${:.3f}\\pm{:.3f}$'.format(snr.value, snr.stderr)

fr = self.fit_res['shots_all']
bv = fr.params
a_sp = bv['A_spurious']
fit_text += '\n\nSpurious Excitations:'
fit_text += '\n$p(e|0) = {:.3f}$'.format(a_sp.value)
if self.options_dict.get('fixed_p01', True) == True:
fit_text += '$\\pm{:.3f}$'.format(a_sp.stderr)
else:
fit_text += ' (fixed)'

b_sp = bv['B_spurious']
fit_text += ' \n$p(g|\\pi) = {:.3f}$'.format(b_sp.value)
if self.options_dict.get('fixed_p10', True) == True:
fit_text += '$\\pm{:.3f}$'.format(b_sp.stderr)
else:
fit_text += ' (fixed)'

if two_dim_data:
offs = self.proc_data_dict['raw_offset']
#fit_text += '\nOffset from raw:\n'
#fit_text += '({:.3f},{:.3f}) {},\n'.format(offs[0], offs[1], eff_voltage_unit)
fit_text += '\n\nRotated by ${:.1f}^\\circ$'.format(offs[2]*180/np.pi)
fit_text += '\n\nRotated by ${:.1f}^\\circ$'.format((offs[2]*180/np.pi)%180)
auto_rot = self.options_dict.get('auto_rotation_angle', True)
fit_text += '(auto)' if auto_rot else '(man.)'
else:
Expand Down
Loading

0 comments on commit 24a0611

Please sign in to comment.