This tutorial covers basic usage of PycQED focusing on running basic experiments using MeasurementControl.
The MeasurementControl is the main Instrument in charge of running any experiment. It takes care of saving the data in a standardized format as well as live plotting of the data during the experiment.
PycQED makes a distinction between soft(ware) controlled measurements and hard(ware) controlled measurements.
In a soft measurement MeasurementControl is in charge of the measurement loop and consecutively sets and gets datapoints. A soft measurement can be 1D, 2D or higher dimensional and also supports adaptive measurements in which the datapoints are determined during the measurement loop.
In a hard measurement the hardware (such as an AWG or a central controller) is in charge of the measurement loop. In this case, the datapoints to be acquired are determined before the experiment starts and are precompiled into the hardware which is then armed and starts acquisition. In a hard measurement MeasurementControl does not take care of the measurement loop but still takes care of the data storage and live plotting of the experiment.
In [1]:
import pycqed as pq
import numpy as np
from pycqed.measurement import measurement_control
from pycqed.measurement.sweep_functions import None_Sweep
import pycqed.measurement.detector_functions as det
from qcodes import station
station = station.Station()
In [2]:
MC = measurement_control.MeasurementControl('MC',live_plot_enabled=True, verbose=True)
MC.station = station
station.add_component(MC)
Out[2]:
The InstrumentMonitor can be used to see the parameters of any instrument connected to the station and updates during the loop initiated by MeasurementControl.
In [3]:
from pycqed.instrument_drivers.virtual_instruments import instrument_monitor as im
IM = im.InstrumentMonitor('IM', station)
station.add_component(IM)
# Link the instrument monitor to the MC so that it gets updated in the loop
MC.instrument_monitor('IM')
IM.update()
In [7]:
IM.update_interval(.1)
IM.update()
In [8]:
from pycqed.instrument_drivers.physical_instruments.dummy_instruments import DummyParHolder
dummy_instrument = DummyParHolder('dummy_instrument')
station.add_component(dummy_instrument)
Out[8]:
In [9]:
MC.soft_avg(15)
MC.persist_mode(True)
MC.set_sweep_function(None_Sweep(sweep_control='hard'))
MC.set_sweep_points(np.linspace(0, 10, 30))
MC.set_detector_function(det.Dummy_Detector_Hard(noise=0.5, delay=.02))
dat = MC.run('dummy_hard')
data_set = dat['dset']
By setting persist_mode = True we can see a copy of the last measurements
In [10]:
MC.set_sweep_function(None_Sweep(sweep_control='hard'))
MC.set_sweep_points(np.linspace(0, 10, 30))
MC.set_detector_function(det.Dummy_Detector_Hard(noise=0.5, delay=.02))
dat2 = MC.run('dummy_hard persistent')
data_set2 = dat2['dset']
In [11]:
dummy_instrument.x(145/134545)
IM.update()
In [12]:
dummy_instrument.delay(.01)
MC.soft_avg(15)
MC.set_sweep_function(dummy_instrument.x)
MC.set_sweep_points(np.linspace(-1,1,30))
dummy_instrument.noise(1)
MC.set_detector_function(dummy_instrument.parabola)
dat = MC.run('1D test')
data_set = dat['dset']
# the second plot will also show the first line
MC.set_sweep_function(dummy_instrument.x)
MC.set_sweep_points(np.linspace(-1,1,30))
dat2= MC.run('1D test-persist')
data_set2 = dat2['dset']
In [11]:
dummy_instrument.delay(.01)
MC.soft_avg(15)
MC.set_sweep_function(dummy_instrument.x)
MC.set_sweep_points(np.linspace(-1,1,30))
MC.set_detector_function(det.Dummy_Detector_Soft())
dat = MC.run('1D test')
data_set = dat['dset']
You can play around a bit with the options in the MC:
In [13]:
MC.persist_mode(True) # Turns on and off persistent plotting
MC.verbose(True)
MC.plotting_interval(.2)
MC.live_plot_enabled(True)
In [13]:
dummy_instrument.delay(.0001)
MC.soft_avg(4)
sweep_pts = np.linspace(-2, 2, 30)
sweep_pts_2D = np.linspace(-2, 2, 5)
MC.set_sweep_function(dummy_instrument.x)
MC.set_sweep_function_2D(dummy_instrument.y)
MC.set_sweep_points(sweep_pts)
MC.set_sweep_points_2D(sweep_pts_2D)
MC.set_detector_function(dummy_instrument.parabola)
dat=MC.run('test', mode='2D')
data_set = dat['dset']
In [14]:
MC.soft_avg(1)
sweep_pts = np.linspace(0, 10, 30)
sweep_pts_2D = np.linspace(0, 10, 30)
MC.set_sweep_function(None_Sweep(sweep_control='hard'))
MC.set_sweep_function_2D(None_Sweep(sweep_control='soft'))
MC.set_sweep_points(sweep_pts)
MC.set_sweep_points_2D(sweep_pts_2D)
MC.set_detector_function(det.Dummy_Detector_Hard(delay=.05, noise=.1))
dat = MC.run('2D_hard', mode='2D')
data_set = dat['dset']
The number of soft_averages determines how many times the experiment will be performed. Only the averaged data is plotted and saved. The number of soft-averages can be set as a parameter of the Measurement Control.
Will first implement it for 1D hard sweeps (easier) and then follow for combinations of hard and soft sweeps.
In [21]:
MC.soft_avg(4)
MC.set_sweep_function(None_Sweep(sweep_control='hard'))
MC.set_sweep_points(np.linspace(0, 10, 30))
MC.set_detector_function(det.Dummy_Detector_Hard(noise=1.5, delay=.02))
dat = MC.run('dummy_hard')
data_set = dat['dset']
In [22]:
MC.soft_avg(10)
sweep_pts = np.linspace(0, 10, 30)
sweep_pts_2D = np.linspace(0, 10, 5)
MC.set_sweep_function(None_Sweep(sweep_control='hard'))
MC.set_sweep_function_2D(None_Sweep(sweep_control='soft'))
MC.set_sweep_points(sweep_pts)
MC.set_sweep_points_2D(sweep_pts_2D)
MC.set_detector_function(det.Dummy_Detector_Hard(noise=1.5, delay=.001))
dat = MC.run('dummy_hard_2D', mode='2D')
data_set = dat['dset']
In [17]:
dummy_instrument.delay(.05)
In [22]:
dummy_instrument.noise(2)
In [24]:
from pycqed.measurement.optimization import nelder_mead
MC.soft_avg(1)
dummy_instrument
MC.set_sweep_functions([dummy_instrument.x, dummy_instrument.y])
MC.set_adaptive_function_parameters({'adaptive_function':nelder_mead,
'x0':[-5,-5], 'initial_step': [2.5, 2.5]})
dummy_instrument.noise(2)
MC.set_detector_function(dummy_instrument.parabola)
dat = MC.run('1D test', mode='adaptive')
data_set = dat['dset']
For a more advanced example of adaptive measurements see "Tutorial: Measurement Control - adaptive sampling".