Message Boards Message Boards

0
|
6841 Views
|
15 Replies
|
2 Total Likes
View groups...
Share
Share this post:

Notation of Wolfram quantum computation framework

Posted 2 years ago
POSTED BY: Doug Beveridge
15 Replies

Due to the number of comments getting quite large on the original discussion (https://community.wolfram.com/groups/-/m/t/2416125 ), they are being split into a separate discussion thread. Please follow the forum rules http://wolfr.am/StaffPicks and stay technical, concise and relevant to Wolfram Technologies. Please keep your further comments relevant to this specific discussion on this thread.

POSTED BY: Moderation Team
Posted 2 years ago

As a side note, the implication on here is that Dirac notation as input to say WQF would only be of interest to the Quantum expert who is familiar with the notations in QM.

I just want to add I am not a quantum expert and got all my info and code for QFT and QPE from courses I did for High School Kids .......

My code on QFT comes from the Qiskit Global Summer School in July 2020 as a two-week intensive summer school for high school kids and undergrades :-

8.Shor's Algorithm I: Understanding Quantum Fourier Transform, Quantum Phase Estimation - Part 2

https://youtu.be/pq2jkfJlLmY?t=224

and associated PDF of lecture is added below.

I could not get on the 2021 Qiskit Global Summer School on Quantum Machine Learning https://qiskit.org/textbook-beta/summer-school/quantum-computing-and-quantum-learning-2021/ because I am not a high school student :)

So all my knowledge and code is high school level :)

Attachments:
POSTED BY: Doug Beveridge
Posted 2 years ago

Here is my final plea to have Dirac Notation added as input to the WQF. This Dirac notation is fundamental to a completer and more versatile WQF. Just trust me on this one, I have been using Dirac Notation in Mathematica using the Quantum AddOn extensively now for more than 5 years and I can attest to its power and versatility, see attached notebook

Attachments:
POSTED BY: Doug Beveridge
Posted 2 years ago

Here is my preliminary comparison between Qiskit, the Wolfram Quantum Framework and the Quantum AddOn.

Attachments:
POSTED BY: Doug Beveridge

Doug, look at our documentations. I suggest you to explore our paclet more before any judgment. Only to mention a few examples for comparison, try to construct Quantum Switch in other packages or qiskit, or construct a controlled-gate with many control and many target.

POSTED BY: Mads Bahrami
Posted 2 years ago

Hi Mads

I will work through that link to the documentation, especially the Algorithms.

I have been using Qiskit and the Quantum add on by José Luis Gómez-Muñoz and Francisco Delgado extensively and and will make a thorough comparison with this paclet .

I just find that the paclet is more Pythonesque (Qiskit) then Woframesque.

I will post a notebook on here of the results of my comparison

For a start I was hoping that the quantum computation framework would include Dirac notation . eg here is the Quantum add on doing fourier transforms and Dirac Notation. ( I will work through the documentation to see how the quantum computation framework does these fourier transforms )

enter image description here

POSTED BY: Doug Beveridge
Posted 2 years ago

I wouldn't say that absence of familiar notation makes it less Woframesque. I've never seen heavy use of notation in standard Wolfram code, unless it's some Domain Specific package, like the Quantum'Computing you mention. The notation is usually considered an add-on, which we can also provide on top of the paclet later. I see more and more "object-oriented" designs around WL nowadays (having high-level objects with a bunch of computed properties), so this one is no exception. For now, the Dirac notation is only part of output formatting, like "Amplitudes" and "Formula" reveal, for example.

You can add formatting, by making states show their simplified formula for example:

FormatValues[QuantumState] =.
Format[qs_QuantumState] := Expand @ FullSimplify @ qs["Formula"]

Then your Fourier example would look like this in the framework:

enter image description here

POSTED BY: Nikolay Murzin
Posted 2 years ago

Python/Qiskit cannot do Quantum computing/dirac notation .Simple. The Quantum AddOn can do Quantum computing /dirac notation very effectively. The new Wolfram Quantum Framework developers obviously made a conscious decision not to go with notation but implement qiskit in Mathematica. Wolfram Quantum Framework gives me no advantage ( in terms of notation ) over qiskit so I will not invest my time in Wolfram Quantum Framework and just stick to Qiskit and the Quantum AddOn ,

I am told I will not be able to implement the Quantum Switch or construct a controlled-gate with many control and many target. But I have the Quantum AddOn which allows me to construct gates in dirac notation so I will give the Wolfram Quantum Framework a miss.

POSTED BY: Doug Beveridge

Your claims are simply wrong, indicating lack of proper knowledge/experience with the Wolfram Quantum Framework, or even with the Wolfram Language. One of the essential features of the Wolfram Language is code readability, even for none expert. Notations you like, are, of course, readable for quantum experts only, to say the least. Note those notations are a minor thing in design, and nothing fundamental at all, only a simple formatting in the input/output, as Nik mentioned.

POSTED BY: Mads Bahrami
Posted 2 years ago

So I can assume you will claim that the adherence to quantum notation (dirac notation ) for inputs and outputs in the Quantum AddOn is just frivolous and that "notations are a minor thing in design, and nothing fundamental at all, only a simple formatting in the input/output" . ( have you even looked at Quantum AddOn, is all that notation just for Quantum experts?

"Notations you like, are, of course, readable for quantum experts only". If you read any books on Quantum Computing or Quantum information, then dirac notation is fundamental to the presentation (see Ike and Mike ) . To say that " Notations is for quantum experts only" is just simply wrong (to quote you . ) . You do you think the principle users of the Quantum Framework are going to be, none expert or aspiring quantum experts.

So why dont you test your "anti notation " stance when the Quantum Framework Course is presented in a few weeks by WolframU . At the beginning of each course we are normally presented with multiple choice Questions about the course ansdour experience with the course material .

Why dont you ask a question like : enter image description here

And I am sure I will be the only one to answer B So you can follow up with

enter image description here

POSTED BY: Doug Beveridge
Posted 2 years ago

I still cannot get my head around your quote "Notations you like, are, of course, readable for quantum experts only, to say the least. "

People coming to the Quantum Framework will have some understanding of Quantum Computing or Quantum information theory. Every textbook or course will present these subjects in terms of Dirac notation
Currently I am working through the proof of the deutsch-jozsa algorithm presented at the Qiskit Global Summer School 2021 and its all in Dirac notation and I need to understand it and I am not a "quantum expert" .

enter image description here

Lets say I want to study the Quantum Fourier Transforms . I go to Mike and Ike page 245 .

enter image description here

and using Mathematica and the Quantum Addonn I can code up that exact equations

enter image description here

If I consult the Qiskit book on Quantum Fourier Transforms at https://qiskit.org/textbook/ch-algorithms/quantum-fourier-transform.html

enter image description here

To say that Dirac Notation is for quantum experts only and it will not be readable to non expert is simply wrong (to quote you )

POSTED BY: Doug Beveridge

I'm pretty sure bra-ket notation will be unfamiliar to those not versed in QM. Then again, one would probably need some QM background to use any of these packages. In any case, I remain unclear on what point(s) you are trying to make. I will add that your escalating annoyance is not clarifying your claims, and it is in common interest for this to return to a civil discussion. I gather the WQF supports traditional QM formatting for output, should one wish to do that. Is your issue that there is not support for bra-let and related infix such as tensor product for inputs?

POSTED BY: Daniel Lichtblau
Posted 2 years ago

"I gather the WQF supports traditional QM formatting for output, should one wish to do that. Is your issue that there is not support for bra-let and related infix such as tensor product for inputs?"

Daniel , thats all I am asking , WQF supports traditional QM formatting for output and I want support for is bra-let and related infix such as tensor product for inputs. This would give WQF much more power . For the non expert they can use WQF as it is but for those that want to use the full extent of the bra-let for input it would make WQF much more powerful . WQF would encompass the Quantum AddOnn.

POSTED BY: Doug Beveridge
Posted 2 years ago

You're right to ask for a familiar notation, because that's how researchers like to write their WL code, by making it as close to textbook examples as possible. But the main point of the framework is to provide a set of high-level abstractions first for people to reason about. You can compare to any other scientific WL framework out there, like Molecule for chemistry, BioSequence for molecular biology, Region for geometry, ClassifierFunction for machine learning, or just any Entity in general for all-around knowledge concepts out there. Quantum Framework has an increasing level of abstractions introducing QuditBasis, QuantumBasis, QuantumState, QuantumOperator, QuantumMeasurementOperator, QuantumMeasurement, QuantumCircuitOperator, and some more. Learning how to input all of those with notation or not requires the same amount of learning curve, and serves different kinds of users. But abstract objects come first, notation sugar later.

POSTED BY: Nikolay Murzin
Posted 2 years ago

If you look at the Quantum AddOn all the Dirac notation is there as input and output , it works perfectly . Any book or course on Quantum Computing or Quantum information is introduced in Dirac notation and quantum computer languages , Qiskit and Q# then code this Dirac notation into code .

The steep learning curve is understanding the Dirac notation . eg If you want to understand Quantum Fourier Transforms you are not going to get that from Qikit , or the Quantum Addon or WQF built in functions .

As stated before you need to look at the QFT formulae eg Mike and Ike enter image description here

"Learning how to input all of those with notation or not requires the same amount of learning curve, and serves different kinds of users. But abstract objects come first, notation sugar later."

The abstract objects are those complicated dirac formulas as in Mike and Ike . If you can code for those then you are mastering QC or QIT, The abstraction is in the notation, master the notation and you master the abstraction.

When I can code like this am I mastering notation or abstraction, I say I am mastering both. So allow these techniques in the WQF

enter image description here

POSTED BY: Doug Beveridge
Reply to this discussion
Community posts can be styled and formatted using the Markdown syntax.
Reply Preview
Attachments
Remove
or Discard

Group Abstract Group Abstract