Some CRUS stats

[vmod]

I've had some time to play with a script and generate some stats about the project.
Attached is an html page (in the familiar greenish format) containing:

Some general stats on the project's progress on conjectures and the work on 1k, 2k, 3k bases

and tables for:
Top-20 proven conjectures by highest CK
Top-20 proven conjectures by largest prime
Top-20 conjectures with 1k remaining by highest CK
Top-20 conjectures with 1k remaining at lowest n by highest weight
Top-20 conjectures with 2ks remaining by highest CK
Top-20 conjectures with 3ks remaining by highest CK
Top-20 conjectures tested to at least n=25K by highest CK
Top-20 conjectures tested to at least n=100K by highest CK

There might be some minor discrepancies (because of bases with k's at different test limits or info on separate pages or powers of 2 bases that were not fully incorporated - I'll see to fix those), however after some generic checking the data look pretty solid.

Perhaps there's more stats and tables that could be generated. For now I think these look thorough and interesting. Enjoy!

[Mathew]

vmod,

I really like the tables. Thanks

[mdettweiler]

Very cool! Do you by chance have a script that you used to generate this? If so, could you send it to me once you're done fiddling with it? I could then set it up on the noprimeleftbehind.net server so it runs and updates automatically, like what we've done for Karsten's crus_tab script.

(I'll bet Gary is cringing right now at the thought of yet another scheduled script relying on the formatting consistency of his web pages... )

[MyDogBuster]

Nice job. Kinda sums up everything very nicely. I especially liked the 1K to n=100K stat. Gives me an idea where I'm at on that.

[gd_barnes]

Excellent work Vmod. A great summary of things. I'll be interested in seeing the power-of-2 bases added.

Max, I don't mind more automated scripts being run from the pages. In the last few weeks, I've attempted to make sure that the formatting stays consistent.

[vmod]

Updated.
1. Bases powers of 2 are now included (flooding the last tables (high ck and high n) and R128 claiming top place for largest Riesel prover-prime).

2. Bases with different k's at different search levels (R6, R31 etc) including those with info on individual reservation pages are now correct. (Well, almost all of them. There are about 3 that it's not possible to pick up the lowest search limit displayed on the separate page. Only one would feature on the tables, so it's rather trivial but to be consistent I have a couple of ideas for a workaround, one should be easy I'll send Gary a PM about this).

3. Added the largest group of consecutive proven bases for Riesel/Sierp (Sierpinski wins).

Max, yes there is a script. Somewhat fragmented at the moment but once finalised it will be a single batch(or bash) that can run and produce the page automatically. Indeed like Karsten's script it depends on the formatting consistency of the pages. Spaces, symbols, brackets, parentheses all matter. As long as the current level of detail is maintained, it should work fine.

I'll go on and give the code some good polishing and once it's ready to go I'll send the script to Max to set it running from the server. If someone has any corrections or suggestions on the terminology I've used or anything else let me know.

[mdettweiler]

Originally Posted by vmod

Updated.
1. Bases powers of 2 are now included (flooding the last tables (high ck and high n) and R128 claiming top place for largest Riesel prover-prime).

2. Bases with different k's at different search levels (R6, R31 etc) including those with info on individual reservation pages are now correct. (Well, almost all of them. There are about 3 that it's not possible to pick up the lowest search limit displayed on the separate page. Only one would feature on the tables, so it's rather trivial but to be consistent I have a couple of ideas for a workaround, one should be easy I'll send Gary a PM about this).

3. Added the largest group of consecutive proven bases for Riesel/Sierp (Sierpinski wins).

Max, yes there is a script. Somewhat fragmented at the moment but once finalised it will be a single batch(or bash) that can run and produce the page automatically. Indeed like Karsten's script it depends on the formatting consistency of the pages. Spaces, symbols, brackets, parentheses all matter. As long as the current level of detail is maintained, it should work fine.

I'll go on and give the code some good polishing and once it's ready to go I'll send the script to Max to set it running from the server. If someone has any corrections or suggestions on the terminology I've used or anything else let me know.

Sounds good. BTW, if the choice is between a batch or bash script, bash would be preferable on this end--the server is running Linux and it can be a little messy to run such things through Wine. Perl would be fine as well.

[MyDogBuster]

What's the status on getting this script into production?

[vmod]

Originally Posted by MyDogBuster

What's the status on getting this script into production?

I was away from home on a (unexpected) trip for the second half of October. As a result many things including this script were on hold.
I'll put some final touches over the weekend, make sure there are no obvious errors and send it to Max early next week.

[MyDogBuster]

I was away from home on a (unexpected) trip for the second half of October. As a result many things including this script were on hold.
I'll put some final touches over the weekend, make sure there are no obvious errors and send it to Max early next week.

Thanks for the update. It's a great script.

[vmod]

I've made a linux adaptation and now the script resides in Max's nplb email inbox. For reference and until it goes live I'm attaching the page with the latest update as of this morning. Quickly cross-checking with the work posted since the previous one I see that everything works and updates as it should. When Max finds time to get the script on the server and provided there are no problems it should run on a schedule and be available from the crus website.

If at any time someone notices anything weird or erroneous on the page post here or PM and I'll see what's wrong. Perhaps something with the script or some webpage formatting inconsistency. As this project progresses, there might be some adjustments or additions on the page but for now I think it's comprehensive, correct (as much as possible) and a nice addition to the rest of the webpages.

PS Since I mentioned formatting; Gary on the Sierp powers of 2 webpage on the 2 odd-n remaining k's part you should remove <br> after 85287 (1.6442M). It's not necessary and causes Sierp 2odd-n to be missing from the last two tables.