Saving Experiment Data to the Cloud¶

Qiskit Experiments is designed to work with Qiskit’s online experiment database, where you can view and share results of experiments you’ve run as ExperimentData objects. This tutorial shows how to save your experimental results to the database. You will need to have qiskit-ibmq-provider installed locally and an account in the Qiskit cloud service. We will use the ibmq_lima backend which is open and available to everyone.

\(T_1\) Experiment¶

Let’s run a \(T_1\) experiment and save the results to the experiment database.

from qiskit_experiments.library.characterization import T1
import numpy as np

t1_delays = np.arange(1e-6, 600e-6, 50e-6)

# Create an experiment for qubit 0,
# setting the unit to microseconds,
# with the specified time intervals
exp = T1(qubit=0, delays=t1_delays)
print(exp.circuits()[0])

     ┌───┐ ░ ┌─────────────────┐ ░ ┌─┐
  q: ┤ X ├─░─┤ Delay(1e-06[s]) ├─░─┤M├
     └───┘ ░ └─────────────────┘ ░ └╥┘
c: 1/═══════════════════════════════╩═
                                    0 

Now we run the experiment. block_for_results() blocks execution until the experiment is complete, then save() is called to save the data to ResultsDB.

from qiskit import IBMQ
IBMQ.load_account()
provider = IBMQ.get_provider(hub="ibm-q", group="open", project="main")
backend = provider.get_backend("ibmq_lima")

t1_expdata = exp.run(backend=backend, shots=1000).block_for_results()
t1_expdata.save()

You can view the experiment online at https://quantum-computing.ibm.com/experiments/10a43cb0-7cb9-41db-ad74-18ea6cf63704

Note that calling save() before the experiment is complete will instantiate an experiment entry in the database, but it will not have complete data. To fix this, you can call save() again once the experiment is done running.

Our \(T_1\) figure and analysis results:

display(t1_expdata.figure(0))
for result in t1_expdata.analysis_results():
    print(result)

../_images/experiment_cloud_service_4_0.png

AnalysisResult
- name: @Parameters_T1Analysis
- value: CurveFitResult:
 - fitting method: least_squares
 - number of sub-models: 1
  * F_exp_decay(x) = amp * exp(-x/tau) + base
 - success: True
 - number of function evals: 6
 - degree of freedom: 9
 - chi-square: 11.482176770249533
 - reduced chi-square: 1.2757974189166148
 - Akaike info crit.: 5.470672047146003
 - Bayesian info crit.: 6.925391996510005
 - init params:
  * amp = 0.9380619380619382
  * tau = 0.00012486915523303855
  * base = 0.006493506493506494
 - fit params:
  * amp = 0.9504789294517044 ± 0.007423697983317722
  * tau = 6.285861999775442e-05 ± 1.5384043570510136e-06
  * base = 0.009541610991605803 ± 0.001460757413231069
 - correlations:
  * (tau, base) = -0.44714123092529423
  * (amp, tau) = -0.16684981921128555
  * (amp, base) = -0.13083060052901557
- quality: good
- device_components: ['Q0']
- verified: False
AnalysisResult
- name: T1
- value: (6.29+/-0.15)e-05
- χ²: 1.2757974189166148
- quality: good
- extra: <1 items>
- device_components: ['Q0']
- verified: False

You can also view the results at the IBM Quantum Experiments pane on the cloud.

By default, the interface displays all experiments you have privilege to see, but this link shows your own experiments. You can change that setting by clicking on the All Experiments dropdown. You can also filter by device, date, provider, and result by clicking on the filter icon.

Individual experiment pages show the plot, and one or more important analysis results, which for the \(T_1\) experiment is the fitted \(T_1\) value.

The metadata field shows experiment metadata included in the ExperimentData object.

You can change the quality and verify/unverify the results upon selection of an analysis result. Quality is an automatic parameter generated by the experiment analysis based on pre-set criteria. The verification field is for a human to determine whether the result is acceptable.

Loading an experiment from the database¶

You can also retrieve the full ExperimentData object from the database service. Let’s load a previous T1 experiment, which we’ve made public by editing the Share level field:

from qiskit_experiments.framework.experiment_data import ExperimentData
service = ExperimentData.get_service_from_backend(backend)
load_expdata = ExperimentData.load("9640736e-d797-4321-b063-d503f8e98571", service)

To display the figure, which is serialized into a string, we need the SVG library:

from IPython.display import SVG
SVG(load_expdata.figure(0).figure)

The analysis results have been retrieved as well:

for result in load_expdata.analysis_results():
    print(result)

AnalysisResult
- name: T1
- value: 0.0001040+/-0.0000028
- χ²: 0.8523786276663019
- quality: good
- extra: <1 items>
- device_components: ['Q0']
- verified: False
AnalysisResult
- name: @Parameters_T1Analysis
- value: CurveFitResult:
- fitting method: least_squares
- number of sub-models: 1
* F_exp_decay(x) = amp * exp(-x/tau) + base
- success: True
- number of function evals: 9
- degree of freedom: 9
- chi-square: 7.671407648996717
- reduced chi-square: 0.8523786276663019
- Akaike info crit.: 0.6311217041870707
- Bayesian info crit.: 2.085841653551072
- init params:
* amp = 0.923076923076923
* tau = 0.00016946294665316433
* base = 0.033466533466533464
- fit params:
* amp = 0.9266620487665083 ± 0.007096409569790425
* tau = 0.00010401411623191737 ± 2.767679521974391e-06
* base = 0.036302726197354626 ± 0.0037184540724124844
- correlations:
* (tau, base) = -0.6740808746060173
* (amp, base) = -0.4231810882291163
* (amp, tau) = 0.09302612202500576
- quality: good
- device_components: ['Q0']
- verified: False

Auto-saving an experiment¶

The auto_save feature automatically saves changes to the ExperimentData object to the cloud service whenever it’s updated.

exp = T1(qubit=0, delays=t1_delays)

t1_expdata = exp.run(backend=backend, shots=1000)
t1_expdata.auto_save = True
t1_expdata.block_for_results()

You can view the experiment online at https://quantum-computing.ibm.com/experiments/cdaff3fa-f621-4915-a4d8-812d05d9a9ca
<ExperimentData[T1], backend: ibmq_lima, status: ExperimentStatus.DONE, experiment_id: cdaff3fa-f621-4915-a4d8-812d05d9a9ca>

Deleting an experiment¶

Both figures and analysis results can be deleted. Note that unless you have auto save on, the update has to be manually saved to the remote database by calling save(). Because there are two analysis results, one for the T1 parameter and one for the curve fitting results, we delete twice.

t1_expdata.delete_figure(0)
t1_expdata.delete_analysis_result(0)
t1_expdata.delete_analysis_result(0)

Are you sure you want to delete the experiment plot? [y/N]: y
Are you sure you want to delete the analysis result? [y/N]: y
Are you sure you want to delete the analysis result? [y/N]: y

The web interface shows that both the figure and analysis result have been deleted:

Saving Experiment Data to the Cloud¶

\(T_1\) Experiment¶

Loading an experiment from the database¶

Auto-saving an experiment¶

Deleting an experiment¶

Tagging and sharing experiments¶

This code is a part of Qiskit