mersenneforum.org  

Go Back   mersenneforum.org > Search Forums

Showing results 1 to 24 of 24
Search took 0.01 seconds.
Search: Posts Made By: tabidots
Forum: Factoring 2019-05-22, 07:55
Replies: 48
Views: 5,766
Posted By tabidots
On second thought, I think they are all probably...

On second thought, I think they are all probably duplicates, because on a typical go-round with these parameters, the program will cycle between the same two sets of qs, and the huge spike in smooths...
Forum: Factoring 2019-05-22, 06:23
Replies: 48
Views: 5,766
Posted By tabidots
When you say sigma=1, you mean sigma=1*mean,...

When you say sigma=1, you mean sigma=1*mean, right?

I think I may have implemented my normal distribution function incorrectly, because I just had something very interesting happen—when I changed...
Forum: Factoring 2019-05-22, 05:25
Replies: 48
Views: 5,766
Posted By tabidots
I just tried with this with the first two B's of...

I just tried with this with the first two B's of the first A of a polynomial that did not pick up any smooths after sieving, and the results were the same—no smooths. That sounds right, since I am...
Forum: Factoring 2019-05-21, 16:45
Replies: 48
Views: 5,766
Posted By tabidots
Wow, your SIQS finds enough relations to find a...

Wow, your SIQS finds enough relations to find a factor after just 14 polynomials? After 14 polynomials, I'm still in the single digits for that 50-digit number!

For now, I have changed the...
Forum: Factoring 2019-05-21, 14:41
Replies: 48
Views: 5,766
Posted By tabidots
All good info, thanks :) I feel like in...

All good info, thanks :)

I feel like in some ways Contini's thesis is a bit of a cruel joke, in that it is rather detailed about the whole algorithm except for the generation of A, which is such...
Forum: Factoring 2019-05-21, 10:13
Replies: 48
Views: 5,766
Posted By tabidots
I am trying different sieve sizes, different...

I am trying different sieve sizes, different methods of finding A, different numbers of used-q's to remove from the q-pool after each polynomial (1, 2, or all), different log-thresholds... But...
Forum: Factoring 2019-05-21, 06:12
Replies: 48
Views: 5,766
Posted By tabidots
I found a library with a weighted-sample function...

I found a library with a weighted-sample function that can take k random samples and an optional weighting function, so I experimented with using a normal distribution as the function, with \mu =...
Forum: Factoring 2019-05-20, 07:57
Replies: 48
Views: 5,766
Posted By tabidots
What is the big picture of what you are doing...

What is the big picture of what you are doing there? This is as far as I could understand (I'm ignoring the multiplier here):


Compute the ideal \hat a = \frac{\sqrt{2n}}{M}
Compute the...
Forum: Factoring 2019-05-19, 02:37
Replies: 48
Views: 5,766
Posted By tabidots
Thanks for the encouragement :) So,...

Thanks for the encouragement :)



So, I'm making progress (if you define progress as different results, not necessarily better ones). I changed the following things:


Use the full...
Forum: Factoring 2019-05-18, 14:05
Replies: 48
Views: 5,766
Posted By tabidots
I've discovered the (a?) reason my SIQS currently...

I've discovered the (a?) reason my SIQS currently struggles to find a factor is that it produces an abundance of duplicate relations. I'll have to rework the A-generating algorithm. Too bad—the Monte...
Forum: Factoring 2019-05-18, 11:53
Replies: 48
Views: 5,766
Posted By tabidots
Here is something really bizarre. I sprinkled...

Here is something really bizarre. I sprinkled some assertions in my MPQS functions to ensure that


For every sieved x, u^2 \equiv v \textrm{mod} n. (That is, (ax + b)^2 \equiv g_{a,b}(x)...
Forum: Factoring 2019-05-18, 10:45
Replies: 48
Views: 5,766
Posted By tabidots
Thanks—I wish I had found that before Contini's...

Thanks—I wish I had found that before Contini's thesis! (By the way, that page preserves Contini's typo—does it affect your program at all?)

Well, I am still tinkering with my SIQS, because I feel...
Forum: Factoring 2019-05-18, 02:40
Replies: 48
Views: 5,766
Posted By tabidots
Ah, scratch everything I just said. I realized...

Ah, scratch everything I just said. I realized that instead of changing the sign of z, I was subtracting it from 1. Man, complicated algebra is really confusing in Lisps.

So no wonder nothing was...
Forum: Factoring 2019-05-18, 01:00
Replies: 48
Views: 5,766
Posted By tabidots
Okay, so my SIQS is still not finding a factor...

Okay, so my SIQS is still not finding a factor even after gathering a number of relations greater than the size of the factor base. Speed aside, that means sieving is okay; as I said, my MPQS works...
Forum: Factoring 2019-05-18, 00:33
Replies: 48
Views: 5,766
Posted By tabidots
I found the typo that was being referred to in...

I found the typo that was being referred to in your superbly-documented code. I actually had been referring to your code previously as well (to the extent that I can read Java), but for some reason I...
Forum: Factoring 2019-05-17, 16:49
Replies: 48
Views: 5,766
Posted By tabidots
Here is some sample test output (just a tiny bit,...

Here is some sample test output (just a tiny bit, obviously, because the data generated by this program is oceanic). Don't know how helpful it will actually be, but just in case:


N =...
Forum: Factoring 2019-05-17, 15:58
Replies: 48
Views: 5,766
Posted By tabidots
Weird... I don't think I changed anything, but...

Weird... I don't think I changed anything, but all my polynomials are testing fine now for b^2 \equiv N (\textrm{mod } a), (though I don't have the original data that generated the polynomial I wrote...
Forum: Factoring 2019-05-17, 13:52
Replies: 48
Views: 5,766
Posted By tabidots
Ah! I seem to be misunderstanding something very...

Ah! I seem to be misunderstanding something very fundamental here.

In MPQS, one side of the congruence was obtained from values of `ax + b`, the "unsquared" polynomial, and the other side was...
Forum: Factoring 2019-05-17, 02:32
Replies: 48
Views: 5,766
Posted By tabidots
Thanks for this suggestion. I haven't gotten...

Thanks for this suggestion. I haven't gotten around to properly debugging the polynomials as jasonp suggested, but I sure hope the fix is as simple as this!

I realized that my threshold was quite...
Forum: Factoring 2019-05-16, 12:13
Replies: 48
Views: 5,766
Posted By tabidots
Perfect, thanks! I will give this a shot later. I...

Perfect, thanks! I will give this a shot later. I had no idea where to start with debugging this.
Forum: Factoring 2019-05-16, 11:34
Replies: 48
Views: 5,766
Posted By tabidots
Ah, right. Yes, I'm doing that, but have not set...

Ah, right. Yes, I'm doing that, but have not set a limit for accepting them. Any value that is not 1 after trial division is saved, and future remainders are checked against it, and merged if they...
Forum: Factoring 2019-05-16, 10:01
Replies: 48
Views: 5,766
Posted By tabidots
Shouldn't that be 32 Bs per A given 6 factors?...

Shouldn't that be 32 Bs per A given 6 factors? (2^(6-1) = 32)

For that 40-digit number I mentioned above, my As have 5 factors, generating 16 bs per A. (2^(5-1) = 16)

Contini's paper says:
Forum: Factoring 2019-05-16, 09:49
Replies: 48
Views: 5,766
Posted By tabidots
For the factors of A, I am constraining them to...

For the factors of A, I am constraining them to between 400 and 4000, following the Python implementation I referenced.


My SIQS is picking up so few smooths per A that even for this 40-digit...
Forum: Factoring 2019-05-16, 02:45
Replies: 48
Views: 5,766
Posted By tabidots
Question SIQS diagnostics

Hello all, this is my first post on the forum.

I recently got interested in factoring while implementing some number theory-related programs in Clojure to teach myself the language. As far as...
Showing results 1 to 24 of 24

 
All times are UTC. The time now is 21:32.


Tue Oct 19 21:32:04 UTC 2021 up 88 days, 16:01, 0 users, load averages: 1.67, 1.67, 1.58

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.