The run_oopsc.py file has been refactored. It should be updated in the Readme file.
If possible, I'd suggest having also one basic example that produces a plot without any additional command line appended command. I.e., typing

python simple_example.py

generates a plot / intended outcome. I wonder whether a configuration / settings file could help to store command line configurations that one is re-using (e.g., 8 -l planar -px 0.03 -pmx 0.03 -i 1000).

For a simulation one need a graph object and a decoder object, which are mutually dependent.
Currently, their relation are defined by the configuration.py module, but is not as straight forward as it could be.
From the CLI, this is no problem. However, if one would like to run a simulation step by step from for example a jupyter notebook, this is quite complicated.

Steps to reproduce

from opensurfacesim.main import initialize, run
code, decoder = initialize((6,6), "toric", "mwpm", enabled_errors=["pauli"])
run(code, decoder, iterations=10, error_rates = {"p_bitflip": 0.1})

Issue

Matplotlib is using Qt5Agg backend. Interactive plotting is disabled.
Drawing 1/1: Initial
Traceback (most recent call last):
  File "*\stuff.py", line 2, in <module>
    code, decoder = initialize((6, 6), "toric", "mwpm", enabled_errors=["pauli"], plotting=True, initial_states=(0, 0))
  File *\opensurfacesim\main.py", line 91, in initialize
    code.initialize(*enabled_errors, **kwargs)
  File "*\opensurfacesim\codes\_template\plot.py", line 35, in initialize
    self.figure.draw_figure("Initial")
  File "*\opensurfacesim\plot.py", line 649, in draw_figure
    self.display(self.figure)
AttributeError: 'Figure' object has no attribute 'display'

Suggested fix

load_interactive_backend(self) should create a self.display for non-inline/non-interactive mpl backends (such as Qt5Agg).

Generate graph with a coordinate on each vertex in stead of just neighbors.
Plot according to coordinate attribute.

Debugging is now performed by added attributes to the decoder classes or the graph classes.
Instead, debug information should be stored in its own class, where all methods related to debugging are self-contained.
Furthermore, debugging is now not available in single simulation

In every iteration all elements of the graph are reset for a new iteration. This is however not necessary on all accounts.

pip install -r requirements.txt gives the following error:

ERROR: Could not find a version that satisfies the requirement pprint (from versions: none) ERROR: No matching distribution found for pprint

Because pprint is a native library, it shouldn't be included in requirements.txt. See also https://stackoverflow.com/questions/63457762/error-could-not-find-a-version-that-satisfies-the-requirement-pprint-from-r-r

Properly reference Blossom5 and Python integration.
Address license issues

I am trying to run the following from the example:
from qsurface.main import initialize, run, BenchmarkDecoder
code, decoder = initialize((6,6), "toric", "mwpm", enabled_errors=["pauli"])
run(code, decoder, iterations=10, error_rates = {"p_bitflip": 0.1})

I am getting the following error:
KeyError Traceback (most recent call last)
in
......
in match_syndromes(self, syndromes, use_blossomv, **kwargs)
60 matching = matching_graph(
61 edges,
---> 62 maxcardinality=self.config["max_cardinality"],
63 num_nodes=len(syndromes),
64 **kwargs,

KeyError: 'max_cardinality'

I have made sure I am on the same version of python as the one in the hub online example, which runs without issues.

=====UPDATE=====

I discovered that the cause of this error is the missing *.ini files when installing with pip.

Steps to reproduce

1. python run_threshold.py uf toric 1000 -l 8 10 12 -p 0.09 0.10 0.11 (directly from README)
2. python run_threshold.py uf toric 1000 -l 8 9 10 12 -p 0.09 0.10 0.11

Issues

1. toric_graph_2D_uf_.csv contains seemingly duplicate rows where the p index is the same, i.e.

L,p,N,succes
8,0.09,1000.0,816.0
8,0.09,2000.0,1588.0
8,0.1,1000.0,732.0
8,0.1,2000.0,1470.0
8,0.11,1000.0,627.0
8,0.11,2000.0,1272.0
10,0.09,0.0,0.0
10,0.09,1000.0,795.0
10,0.1,0.0,0.0
10,0.1,1000.0,710.0
10,0.11,0.0,0.0
10,0.11,1000.0,659.0
12,0.09,0.0,0.0
12,0.09,1000.0,819.0
12,0.1,0.0,0.0
12,0.1,1000.0,721.0
12,0.11,0.0,0.0
12,0.11,1000.0,638.0

2. Exception: cannot handle a non-unique multi-index!

This is because sim.py appends decoder data to a pandas MultiIndex through a floating point, introducing slight variations depending on how this value is read.

Proposed solution
Write the p column as data, not as an index. Replace with labels that can be uniquely defined, e.g. integers.

Adding a My Binder badge in the Readme file will allow prospective users to try out the software without installing anything locally. I tried it and it works, thanks to the new package structure, with the requirements.txt file in the topmost folder.

Pasting this in README.md should suffice:

(this is how it looks like: [![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/watermarkhu/OpenSurfaceSim/master))

To be most effective, the My Binder link could point directly to a tutorial, e.g. in the form of a Jupyter notebook, or to a script (using Jupyter Labs as My Binder default). Happy to help in this with PR or feedback along the way.

Move plot parameters to a .plotconfig file

CI should be added to check for compatibility

Use pytest to setup some proper codebase testing.

Previously, Patch3D objects wouldn't update colors, to which a workaround was made to plot several versions of the same object in the same location and to select one as the visible object. This bug was tracked in matplotlib/matplotlib#3370.

A recent pull request fixes this bug matplotlib/matplotlib#18189. The workaround can thus be removed.

Now all decorators related to debugging are in the info module. Each decorator function/class should in stead be in their respective modules.

Hi, first or all, thank you for making this amazing repo.
I really like the plotting feature that allows me to understand what is happening.
I have three issues/questions, and I would really appreciate if you could solve these, especially the first one.

1. Plotting unionfind decoder

I'm running the following in jupyter notebook.

%matplotlib qt
from qsurface.main import initialize, run
code, decoder = initialize((5,5), "toric", "unionfind", enabled_errors=["pauli"], plotting=True, initial_states=(0,0))
run(code, decoder, error_rates = {"p_bitflip": 0.05, "p_phaseflip": 0.05}, decode_initial=False)

For the first three plots (initialization, error applied, ancilla qubits measured, everything is fine.
But when I hit next, another window pops up, and the grid doesn't show anymore.
It looks like this:

Could you please solve this, so that the grid shows up, just like when I use the mwpm decoder?

2. Warning when I press "previous" in interactive plot

I have the following message:
Traceback (most recent call last):
File "C:\Users\lg\anaconda3\lib\site-packages\matplotlib\cbook_init_.py", line 270, in process
func(*args, **kwargs)
File "C:\Users\lg\anaconda3\lib\site-packages\matplotlib\widgets.py", line 225, in
return self._observers.connect('clicked', lambda event: func(event))
File "C:\Users\lg\anaconda3\lib\site-packages\qsurface\plot.py", line 695, in _draw_prev
return self._draw_from_history(self.history_iter > 1, -1, **kwargs)
File "C:\Users\lg\anaconda3\lib\site-packages\qsurface\plot.py", line 682, in _draw_from_history
self.draw_figure(**kwargs)
File "C:\Users\lg\anaconda3\lib\site-packages\qsurface\plot.py", line 648, in draw_figure
self.focus()
File "C:\Users\lg\anaconda3\lib\site-packages\qsurface\plot.py", line 471, in focus
output = self.blocking_input(self.params.blocking_wait)
File "C:\Users\lg\anaconda3\lib\site-packages\qsurface\plot.py", line 159, in call
return super().call(n=1, timeout=timeout)
File "C:\Users\lg\anaconda3\lib\site-packages\matplotlib\blocking_input.py", line 93, in call
self.fig.canvas.start_event_loop(timeout=timeout)
File "C:\Users\lg\anaconda3\lib\site-packages\matplotlib\backends\backend_qt5.py", line 424, in start_event_loop
raise RuntimeError("Event loop already running")
RuntimeError: Event loop already running

3. Question: What is info / bitflip / phaseflip in this image?
   

Again, I really appreciate this repo, and I hope you reply!

Graph objects should be separated from the graph base elements (GBE) such as Vertex and Edge.
We should define a base graph class and base GBE classes that are stripped from decoder specific attributes.
Every decoder should add decoder specific methods and attributes through inheritance, defined in their modules_decoder folder.

This allows graph object attributes and methods to be added at the decoder level in stead of the main level.

The two most common docstring formats are Numpy-style and Google-style (more compact). They both contain standard keywords for arguments, returned objects and the possibility to annotate the signature with type hints or within the docstring itself.

This is particularly helpful to automatically build the API-doc as part of the documentation with Sphinx's autodoc's feature.

Numpydoc helps ensuring this happens, https://numpydoc.readthedocs.io/en/latest/format.html#.

Happy to help on this with contributions or feedback.

With the updated codebase, existing tests are not valid anymore.

In the interactive 3D plot, erasure errors cannot be enabled.

Running multiprocessed codes on larger lattices (for both union-find and MWPM) results in a recursion depth error for the pickled input arguments, i.e. code and decoder objects. For faulty measurements, this error happens for lattices as small as 4x4.
This happens on Windows. While it works perfectly in the Unix environment.
Steps to reproduce

from qsurface.main import initialize, run_multiprocess

code, decoder = initialize((4, 4), "toric", "unionfind", enabled_errors=["pauli"], faulty_measurements=True)
if __name__ == "__main__":
    run_multiprocess(code, decoder, iterations=100, error_rates={"p_bitflip": 0.01})

Error

Traceback (most recent call last):
  File "c:\Users\*\Desktop\test.py", line 5, in <module>
    run_multiprocess(code, decoder, iterations=100, error_rates={"p_bitflip": 0.01})
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\site-packages\qsurface\main.py", line 276, in run_multiprocess
    worker.start()
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\multiprocessing\process.py", line 121, in start
    self._popen = self._Popen(self)
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\multiprocessing\context.py", line 224, in _Popen
    return _default_context.get_context().Process._Popen(process_obj)
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\multiprocessing\context.py", line 327, in _Popen
    return Popen(process_obj)
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\multiprocessing\popen_spawn_win32.py", line 93, in __init__
    reduction.dump(process_obj, to_child)
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\multiprocessing\reduction.py", line 60, in dump
    ForkingPickler(file, protocol).dump(obj)
RecursionError: maximum recursion depth exceeded while calling a Python object
PS C:\Users\*> Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\multiprocessing\spawn.py", line 116, in spawn_main
    exitcode = _main(fd, parent_sentinel)
  File "C:\Users\*\AppData\Local\Programs\Python\Python310\lib\multiprocessing\spawn.py", line 126, in _main
    self = reduction.pickle.load(from_parent)
EOFError: Ran out of input

Hello,
I am a new starter and trying to produce the simplest case for the surface code by using qsurface.

I will be appreciate, if you can check the code:
I will put no error to the simulation
1)I want to create the surface code for distance 3
2)I want to initialize all the data qubits in 0 state
3)I want to put X and Z stabilizers
4)Then I want to measure all the X stabilizers. In this case they will randomly change and commute with Z stabilizers and Z stabilizers will not change
With this scenario, I want to see how mwpm is working

As code:
code, decoder = initialize((3,3), "planar", "mwpm", enabled_errors=[], plotting=True, initial_states=(0,0))
run(code, decoder, error_rates = {}, decode_initial=True) #"p_bitflip": 0.1, "p_phaseflip": 0.1

But I could see nothing. Could you help me for my first code?

Hi,
could you provide some basic hints on how to produce an animations from a jupyter notebook like the ones you show in the example?
Thanks a lot

Improve next plot button/enter behavior.
Catch closed window error.