Some CRUS stats

[rogue]

Those problems are fixed. I've updated crus-stats.htm and vstats.zip file on my website.

The decimal length issue was one of rounding. The incorrect values in the "searched to 25K" table were due to using ">" instead of ">=". I've add a "Top-20 Conjectures Tested to at least n=100K by Largest Number of Remaining k" table and put the various 1k/2k/3k tables next to one another at the bottom.

I also cleaned up the 22_build_tables.txt script more to eliminate a lot of duplicate code and use it to build the list of links for the tables rather than hard-code them in 42_final_html.txt. This will make it easier to not only add new tables, but also re-arrange them into a different order.

I noticed that the 2k and 3k tables have a little overlap. Maybe we need to add a few more entries in them to eliminate the overlap.

[rogue]

I've updated the scripts again. I fixed one mistake and created two new tables. One is labeled "Status Summary by Remaining k" and replaces the "<xx> of Riesel conjectures with <xx> remaining have been tested to at least <xx>". The other is called "Top-20 Conjectures with 1k Remaining by Lowest Weight".

Note that my web-page doesn't necessarily reflect current statuses. I was using two day old pages.

[gd_barnes]

Originally Posted by rogue

I've updated the scripts again. I fixed one mistake and created two new tables. One is labeled "Status Summary by Remaining k" and replaces the "<xx> of Riesel conjectures with <xx> remaining have been tested to at least <xx>". The other is called "Top-20 Conjectures with 1k Remaining by Lowest Weight".

Note that my web-page doesn't necessarily reflect current statuses. I was using two day old pages.

Ah, much better. I've now had time for a detailed check. Here are mostly nits with a couple of more issues that I found:

1. The 1st table name has "conjecture" misspelled. It is missing a "t".

2. The 3rd table has headings of "Riesel" and "Limits". "Limits" should be "Sierpinski". Perhaps you want to change both and have them show "Riesel Limits" and "Sierpinski Limits".

3. The decimal length is incorrect on 2 tables..."Top-20 unproven conjectures by tested n" and "Top-20 unproven conjectures by decimal length". It needs to reflect the testing size of the smallest k remaining. Here is an example for Riesel base 2 where the smallest k remaining is k=2293: The decimal length would be [2000000*log(2)+log(2293)+1]=602064 (vs. 602060). This is more important on the 1k conjectures in that table.

4. On tables with a heading of "the k", please change that to "k remaining" or "k's remaining" or "remaining k" or "remaining k's"...something to make the tables mostly consistent.

5. Please put a space between k's separated by a comma for clarity as in "1597, 36772".

6. For consistenty with the 1k and 2k remaining conjectures, please put the table for "conjectures with 3k remaining by highest conjectured k" above the table for "conjectures with 3k remaining by lowest conjectured k".

7. This is a question for you that you can decide upon: Should the base show as reserved when only part of the k's are reserved? This sticks out the most on both sides of base 25 and Riesel base 256 where a small minority of the k's are reserved by other project(s). It gives the impression that the entire base is reserved. I would somewhat prefer that such bases show as unreserved on your page, which forces the user to look at the actual reservations page(s) to see which k's are available.

Thanks for your effort on this.


Gary

[rogue]

Originally Posted by gd_barnes

3. The decimal length is incorrect on 2 tables..."Top-20 unproven conjectures by tested n" and "Top-20 unproven conjectures by decimal length". It needs to reflect the testing size of the smallest k remaining. Here is an example for Riesel base 2 where the smallest k remaining is k=2293: The decimal length would be [2000000*log(2)+log(2293)+1]=602064 (vs. 602060). This is more important on the 1k conjectures in that table.

The problem I have with this one is that I don't "scrape" the list of k from the reservation page, thus I don't know the largest k, thus I cannot compute the exact decimal length. I changed the column label to "Min Decimal Length" for those two tables to indicate that any prime found will be larger than that length.

4. On tables with a heading of "the k", please change that to "k remaining" or "k's remaining" or "remaining k" or "remaining k's"...something to make the tables mostly consistent.

I didn't like "k remaining" or "remaining k" because one could either interpret that as a number of k's or an actual k value (or list of k's). Maybe someone can think of a better label.

7. This is a question for you that you can decide upon: Should the base show as reserved when only part of the k's are reserved? This sticks out the most on both sides of base 25 and Riesel base 256 where a small minority of the k's are reserved by other project(s). It gives the impression that the entire base is reserved. I would somewhat prefer that such bases show as unreserved on your page, which forces the user to look at the actual reservations page(s) to see which k's are available.

Similar to item #3, I don't read the list to determine if some or all are reserved. It just indicates that at least one k is reserved. I don't know what the effort is to indicate partial vs. full reservation vs. reserved by project vs reserved by individual.

The other items have been addressed and I've posted the updates.

[rogue]

I updated the script to handle conjectures with mixed reservation status. This required the addition of a color legend.

[gd_barnes]

Originally Posted by rogue

The problem I have with this one is that I don't "scrape" the list of k from the reservation page, thus I don't know the largest k, thus I cannot compute the exact decimal length. I changed the column label to "Min Decimal Length" for those two tables to indicate that any prime found will be larger than that length.



I didn't like "k remaining" or "remaining k" because one could either interpret that as a number of k's or an actual k value (or list of k's). Maybe someone can think of a better label.



Similar to item #3, I don't read the list to determine if some or all are reserved. It just indicates that at least one k is reserved. I don't know what the effort is to indicate partial vs. full reservation vs. reserved by project vs reserved by individual.

The other items have been addressed and I've posted the updates.

OK on the first one.

On the second one, I would prefer it to be like the current pages. That is "k remaining or k's remaining" on tables where specific k's are being shown and "number of k's remaining" or "# of k's remaining" on other tables to indicate how many k's are remaining. That way, the two different meanings can be differentiated. I suppose just simply "k" or "k's" would be OK for the former meaning. Of course with "k" for the 1k remaining and "k's" for the >1k remaining bases. Maybe that would fit better with what you think is appropriate. I just don't like "the k".

On the 3rd one, nice work on the pink and yellow for partially and fully reserved bases. Good idea. I was going to be more insistent that partially reserved bases not be shown as reserved but your idea is better.

[rogue]

I changed "The k" to "Remaining k". The zip and html have been updated.

One other thing I was thinking was that maybe it the three summary tables at the top should be moved to a different page, then what remains could be placed on a "Top 20" page.

[rogue]

I had some fun with the stats page. I've done the following:

1) Made the tables sortable by clicking on the column header.
2) Expanded the tables to contain all rows matching the criteria.
3) Since tables can be sorted, many of them contain duplicate data, thus ones with duplicate data were either removed or merged with other tables.

I could probably go further remove the 1k/2k/3k conjectures from the first table thus having 1k/2k/3k/>3k unproven conjectures, but I think it is better to see them combined as the top rows could be considered "most wanted".

Thoughts? Opinions?

[Puzzle-Peter]

I seems the columns are centered vertically which looks a bit weird imho. Could you top-align them so the headers are at the same height?

Btw great work!

[rogue]

Originally Posted by Puzzle-Peter

I seems the columns are centered vertically which looks a bit weird imho. Could you top-align them so the headers are at the same height?

Btw great work!

Fixed. I was using "align" as opposed to "valign".

[gd_barnes]

On your original page, to differenciate between the two different kinds of "remaining k", please change the other table headings to show "# of remaining k".

Your latest page would need to be a separate page from the original page.