Bring Tutorial and Example Notebooks Back Up-to-Date

[coreyostrove]

A number of our tutorial notebooks have fallen out of date with respect to the most recent feature sets in pygsti and are in need of either a refresh, or in some cases the creation of new tutorial and example notebooks. Examples of subjects that might need some TLC:

• Experiment design construction tutorials (for using/explaining new germ, fiducial and FPR tools). Maybe a new tutorial on using the fisher information to evaluate experiment designs.
• Fitting error rates models to volumetric benchmarking data vis-a-vis this paper (https://arxiv.org/pdf/2305.08796.pdf)
• New weak simulation capabilities, Kraus operator representations and Qibo integration(?).
• Serialization using MongoDB
• Single parameter wildcard models
• Parity POVM example
• Specifying reduced error generator models using new coefficient block interface.
• etc.

We'll also want to keep on top of making new tutorial and example notebooks as we integrate new capabilities in preparation for the next release.

---

Editing to keep track of progress of subfixes. Checked boxes here means the fix is pushed up to the bugfix-example-updates branch.

• Experiment design construction tutorials (for using/explaining new germ, fiducial and FPR tools). Maybe a new tutorial on using the fisher information to evaluate experiment designs. @coreyostrove?
• Fitting error rates models to volumetric benchmarking data vis-a-vis this paper (https://arxiv.org/pdf/2305.08796.pdf) @dhothem?
• Serialization using MongoDB @enielse?
• Single parameter wildcard models @sserita
• Parity POVM example @sserita (should also solve Flexible measurements #282)
• Specifying reduced error generator models using new coefficient block interface. @sserita (this would probably see an update with Better interface to describe HSCA error generators #277, but I can provide something for now)
• 00-Protocols tutorial issues
• 01-Esssential-Objects tutorial issues
• 02-Using-Essential-Objects tutorial issues
• CustomOperator tutorial issues
• Add procedurally generated error bars to error bar tutorial
  1Q_error_bars_demo.ipynb.zip

I am specifically keeping the "New weak simulation capabilities, Kraus operator representations and Qibo integration(?)" out of Jupyter notebooks for now. I consider this an alpha feature and purposefully do not want to advertise this yet.

[nemill]

A potential tutorial issue, very minor, in "01-Esssential-Objects"

The note above a given cell says:
Another thing to note is that DataSet objects are "sparse" in that 0-counts are not typically stored:

Cell code:
c = Circuit([('Gxpi2',0)], line_labels=(0,1))
print("No 01 or 11 outcomes here: ",ds_fake[c])
for outlbl, cnt in ds_fake[c].counts.items():
print("Item: ",outlbl, cnt) # Note: this loop never loops over 01 or 11!

Output:
No 01 or 11 outcomes here: {('00',): 45.0, ('01',): 0.0, ('10',): 55.0, ('11',): 0.0}
Item: ('00',) 45.0
Item: ('01',) 0.0
Item: ('10',) 55.0
Item: ('11',) 0.0

Issue/Expected Result:
It seems from the comment that the 01 and 11 outcomes should not be printing, yet they are. I'm unsure if this is an artifact of my specific environment, or of modifications made in recent pyGSTi versions.

Environment:
pyGSTi version 0.9.11.1
python version 3.10.9
OS - Windows 11

[nemill]

Minor typos in tutorial 02-Using-Essential-Objects

(1) Randomized benchmarking section:
"More more details on using these protocols..."

(2) Data Comparison Tests section:
The output of the following code:
print("\nCompare two inconsistent DataSets (generated from different model)")
comparator_A1_B = pygsti.data.DataComparator([dsA1,dsB])
comparator_A1_B.run(significance=0.05)

produces the output lines:
Compare two inconsistent DataSets (generated from different model)
The data are INCONSISTENT at 5.00% significance.

• Details:
  • The aggregate log-_likelihood ratio test is significant at 15.45 standard deviations.
  • The aggregate log-_likelihood ratio test standard deviations signficance threshold is 2.02

The "log-_likelihood" notation seems unintended - either "-" or "_". Presumably this extends to the data comparator output functions as a whole, not just to this tutorial.

[nemill]

Typo and Deprecated Code Warnings in "00-Protocols" tutorial

Typo:
In Randomized Benchmarking text, missing parenthesis after the word tutorial in this sentence: "To create the experiment design, the user specifies a QubitProcessorSpec object that describes the quantum processor (see the ProcessorSpec tutorial, ..."

Various warnings from deprecated code:
Environment:
pyGSTi version 0.9.11.1
python version 3.10.9
OS - Windows 11

Gate Set Tomography:

from pygsti.modelpacks import smq1Q_XYI

# get experiment design
exp_design = smq1Q_XYI.create_gst_experiment_design(max_max_length=32)  

# write an empty data object (creates a template to fill in)
pygsti.io.write_empty_protocol_data('tutorial_files/test_gst_dir', exp_design, clobber_ok=True)

# fill in the template with simulated data (you would run the experiment and use actual data)
pygsti.io.fill_in_empty_dataset_with_fake_data(
    "tutorial_files/test_gst_dir/data/dataset.txt",
    smq1Q_XYI.target_model().depolarize(op_noise=0.01, spam_noise=0.001),
    num_samples=1000, seed=1234)

# load the data object back in, now with the experimental data
data = pygsti.io.read_data_from_dir('tutorial_files/test_gst_dir')

# run the GST protocol on the data 
results = pygsti.protocols.StandardGST().run(data)

# create a report
report = pygsti.report.construct_standard_report(
    results, title="GST Overview Tutorial Example Report")
report.write_html("tutorial_files/gettingStartedReport")

Warning Message
C:\Users\nemill\pyGSTi\pygsti\modelmembers\operations\lindbladerrorgen.py:1691: UserWarning: Using 'CPTP' as a Lindblad type is deprecated, and you should now use 'CPTPLND'
_warnings.warn("Using 'CPTP' as a Lindblad type is deprecated, and you should now use 'CPTPLND'")

RPE:
Code Block

# An experiment design
from pygsti.modelpacks import smq1Q_Xpi2_rpe, smq1Q_XYI
exp_design = smq1Q_Xpi2_rpe.create_rpe_experiment_design(max_max_length=64)

# write an empty data object (creates a template to fill in)
pygsti.io.write_empty_protocol_data(exp_design, 'tutorial_files/test_rpe_dir', clobber_ok=True)

# fill in the template with simulated data (you would run the experiment and use actual data)
pygsti.io.fill_in_empty_dataset_with_fake_data(
    "tutorial_files/test_rpe_dir/data/dataset.txt",
    smq1Q_XYI.target_model().depolarize(op_noise=0.01, spam_noise=0.1),
    num_samples=1000, seed=1234)

# read the data object back in, now with the experimental data
data = pygsti.io.read_data_from_dir('tutorial_files/test_rpe_dir/')

# Run the RPE Protocol
results = pygsti.protocols.rpe.RobustPhaseEstimation().run(data)

print(results.angle_estimate)

Warning Message
C:\Users\nemill\pyGSTi\pygsti\io\writers.py:508: UserWarning:
write_empty_protocol_data has recently changed its signature - it looks like you need to swap the first two arguments. Continuing using the old signature...

[nemill]

In "Custom Operator" tutorial (Tutorials/objects/advanced/CustomOperator.ipynb):
Broken link in the cell:
"We'll add a MyXPi2Operator instance as the "Gx" gate in pyGSTi's standard {Idle,𝑋(𝜋/2),𝑌(𝜋/2)} model (see the standard modules tutorial for more information on standard models)."

The link attempts to direct to Tutorials/Objects/Advanced/StandardModules.ipynb which no longer exists or has been moved. This could have possibly been renamed as ModelPacks.ipynb in the same directory.

[kmrudin]

(H/T @nemill) In "Custom Operator" tutorial (Tutorials/objects/advanced/CustomOperator.ipynb), there is a sign error in the definition of the quantity c (In[1], line 25). We can also make both the b and c quantities easier to parse via trig identities. We should replace In[1], lines 25 and 26 (in the from_vector method definition) with

b = a*np.sin(2*theta)
c = a*np.cos(2*theta)

[sserita]

Hi all, thanks for the bug reports here!

Edit: This used to be a checklist but I moved to the original comment so that we can keep track of the subfixes at a glance from the Issues page.

[sserita]

I added wildcard and procedural error bar tutorials in 49fc6c6, and am deeming the other items as still experimental enough that we can handle them at a later time. Given the impending slew of changes to tutorials, I'm closing this issue and we will inevitably open another one of these in the future.