For me, I am very sure it is solved. If you have interest, first download the problem, run it. Then, read my paper and think, then, you may also be sure.
How to use the program
1. I believe that most of people who download my program would be professionals. So I please you leave your contacting message and welcome your opinions if you download my program. You can leave your message here or to my email: [email protected]. Thanks a lot.
2. This program is an informal one, also it is not the quickest one. But it includes my algorithm, also it can work correctly and works very well. No fails for it.
3. How to use: if you have a 0-1 matrix standing for a simple undirected graph with n vertices which has at least a Hamilton path from vertex 0 to vertex n-1, you only press the “ReadMatrix” menu item to read and calculate it, then you press the “Write the result” menu item, to write the result in a new file, you can get a Hamilton path from vertex 0 to vertex n-1 in the new file.
4. How to use: if you have an edges matrix standing for a simple undirected graph with n vertices which has at least a Hamilton path from vertex 1 to vertex n, you only press the “ReadEdges” menu item to read and calculate it, then you press the “Write the result” menu item, to write the result in a new file, you can get a Hamilton path from vertex 1 to vertex n in the new file. If without such a path, you get a message “no...”. The input file format: each row: 1,3 or 1 3. It means that an edge from vertex 1 to vertex 3.
5. The maximum degree is 3. Though I am very sure my algorithm can calculate any degree of undirected graphs, but this program not. The maximum vertex number is 3000, due to that the PC memory is limited.
6. I would like to thank Professor Alexander Chernosvitov very much. He and his one student take a long time to write a program (different from mine) to implement my algorithm and he give me and my work a good comment (see the web codeproject.com and researchgate.net). Mr. Wang, xiaolong also. Before them, no body trust me. Some not smart enough editors and reviewers reject me just on this logic: for such a hard problem, Lizhi Du is not a famous man, so he cannot solve it. Some editor or reviewer does not use his or her brain, say: your paper is apparently wrong, or, your paper cannot be understood. “apparently wrong”, funny! I study it for a lot of years, apparently wrong! If a reviewer is really powerful and use his brain and cost his time, he surely can understand my paper. If you think I am wrong, tell me where is wrong, then I explain to you that is not wrong. If you think my paper cannot be understood, tell me where cannot be understood, I explain to you. In my paper, in the Remarks, I told how to understand my algorithm and proof. I think it is very clear.
7. I studied this problem for a lot of years. I put a lot of versions of my papers on arxiv. Though the former versions have this or that problems, I am very sure the newest version of my paper is the final version and it is surely correct. It may contains some little bugs due to my English expression, but this does not affect the correctness and I can explain or revise the little bugs easily.
8. Surely I think I have proved NP=P and have solved the problem NP vs. P.
9. Thank you for you pay your attention and time on my algorithm!
1, I think that if one does not want to pay time on the paper, paying time on the code is meaningless to me, or cannot understand the paper still means cannot understand the code.
2, thank you for your answer. You are right. I also met the problems what you said. I still cannot compile his code. I seems having told him. You English expression is fine, so can you directly talk to him? Thank you!
https://www.researchgate.net/profile/Alexander_Chernosvitov
After this discuss reading I see the millenium problem is not solved. I believe a program can't prove solution of any theoretical problem.
The complexity of a program and the complexity of an algorithm are different things. The program needs to calculate, get computer memory for calculation, copy data, etc. So how calculate or measure the complexity of a program?
A program realizes an algorithm. Why measure the complexity of the program, if we can measure the complexity of the algorithm. Theoretically.
To Lizhi Du: We have to be correct.
To Peter Breuer: I glad to discuss it with you. Thanks.
to Peter Breuer:
I did not read his code. I will be pleased to see you discuss directly with him.
He said his one student cost one month to develop this code. But I think, one month is not enough for write code, debug and test.
I am not sure his code is right. but should be near to.
also my code is not formal, but I am sure my code can calculate any hard graph.
Thank you for your discussion.
To Peter Breuer:
Thank you. You are right.
two reasons I do not open my code:
1, my code can calculate 3SAT with 100% correctness. I use 13 vertices to mimic a clause, 2 vertices to mimic a variable. In big computer, this means any NP can be done and RSA code can be destroyed. This can bring trouble to me.
2, I have my personal style on my code: no comments, terribly chaotic, also it contains a lot of garbage, because I revised it from the oldest version a lot of years ago. A lot of code are useless, but not deleted, still run. only I can read it. Others cannot.
interesting, I met a judge.
My goal: let my paper be understood. If who can helpful to me for this goal, I have interest, if not , I do not have interest.
I do not want to open my code. I said it again and again and again. Why ask me for this again and again and again.
If not, I may be judeged .....?
About my algorithm, first depend on my writing in the paper , secondly depend on one's intelligence.
I do not know C#, I did not read his code. I do not have interest to discuss his code with people. I tell you again and again to discuss with him direcly if you terribly like to do so. to discuss it is meaningless to me.
But if his code even only partly implement my algorithm, and he tested it a lot, the results are very good. This can not show something, according to you?
I think his code has meanings. So if you really have interest, do something based on his code, you can get a lot. Why you always nitpick other people? What you said is only little detail, not important. You like to nitpick such little things.
Ok, stop meaningless quarrel.
By the way, to all:
What I want is: test the exe, read the paper, then discuss them with me. If you do not like to do so, I can not compel any one.
Thanks a lot.
Dear Peter Breuer:
You are a very persistent man.
Also, as you said, all your words means that you are " more software engineer than mathematician".
My only answer is:
1, you can directly talk to him. And if you really like to pay your time to understand his C# code, run it and debug it, then you can judge it utterly, but you did not.
2,if his code is in C, C++, VC++, I like to utterly to run it and comment it, but for C#, I do not like to do this. I really can cost some time to know C# and then understand his code, but I do not think this is meaningfull to me. So I do not do.
Thank you fou your all comments.
Your logic!
for 1, 2, you like to do it, I do not like to do it. So I said: if you like to do it, you do it with him. if you do not like, I never never like to compel you to do. I am not as you, use your un-logical logic to compel other body to do what you said, often.
for the last question: a tree, its all branchs (routines) are n!. to delete wrong routines, or only choose (find) possible correct routines, this is very logical, very easy, why you cannot understand this?
To: Peter Breuer:
a long time ago, you asked me to open my code. I sai I did not do. Then you asked again and again.
This time, you still again and again asked for this. I still gave a certain answer: no way.
Also you always use this logic: if you do not open your code, then I judge you are.., and your work is .....
Any noamal man who read all of your comments , can immediately get the result: who needs some special psychological intervention.
To: peter:
You are reasonable in some way, so I may open my code some time later, but not now. So now stop discussing code!
Now I only concern the exe and the paper.
everybody can download the exe and paper. If you think the exe cannot calculate some graphs correctly (I think this is not ppossible), tell me.
and any questions about the paper's understanding (where in the paper?), I like to answer and discuss.
Sorry I do not discuss other things.
Thank you, everybody.
I'm sorry, but I agree with Peter. Nobody with a marginal understanding of computer security will run a binary of unknown provenance without extreme caution. Moreover, a significant fraction of the research community does not run Windows (I develop on Macs and do systematic testing on Linux clusters).
Also, the result of any such run would be of, at best, marginal value. Black box testing, e.g. fuzzing, is a useful technique to find some kinds of errors like buffer overflows, but it is unlikely to find corner cases for graph problems. Note that current CDCL solvers regularly solve SAT instances with millions of variables without anyone claiming this as evidence for P=NP. Also note that many high-performance SAT engines are Open Source, which is one of the reasons for their success both in solving instances, but in particular in communicating the ideas that make them work efficiently.
There have been many proposals for showing P=NP, including, IIRC, 9 previous ones by you. All but this last one have been refuted so far. Extraordinary claims need extraordinary evidence.
To:
Open your eyes.
I have two things here: exe and a paper. The paper is the main, the exe is the additional.
The exe can show something, if you can run it. If you do not want to run, I do not mind. Go ahead.
You do not want to run, does not means all others are as you.
Also, read the paper,use your brain to think. You may get something. Also if you do not want to read it, go out, I do not mind. Never all other people are as you.
This is an algorithm, 100% guarantee. CDCL is not.
Apparently, here only wecome two kinds of people: can run the exe or can read the paper. Others can leave here quickly.
Anybody interested in having a look at some pretty clear code should start reading the file LizhiDu.cs in Alexander Chernosvitov's project. That's where we can see the big picture as well as details -- as they were understood by someone else. I don't think any other files are needed.
There is nothing particularly Oriented Object nor C## to know, only small syntax variants departing from simple C. Read it as pseudo-code.
It's not fantastic code either, but it is very useful as a starting point. My current difficulty with it now is to understand how it keeps track of some breaks in its global arrays (the adjacency matrix and the additional breaks[,] array).
Peter, the code in SubSet.cs is not used anywhere in Alexander's project. Forget about it! You cannot reproach Alexander to leave some corner of his project under construction, as long as there is no use of it through the main source tree.
Apparently, Norbert Blum is convinced of a negative solution to the question. See "A Solution of the P versus NP Problem" up on arXiv.
https://arxiv.org/abs/1708.03486
To: Stephan Schulz
Thank you!
1, He seems to be a famous guy. Experts pay notes on him. I am not famous, so experts do not pay notes on my job.
2, My job is absolutely correct.
3, i.e., “experts” pay their time and notes on a wrong job, but do not want to pay their notes on a correct job. Stupid “experts” often judge something by pre-assumption, not by carefully reading, thinking, and correct understanding.
4, I may consider to open my source code, let them see.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python