The New Sudoku Players' Forum

[champagne]

Facing problems to match the current situation and the past, I decided to redo the SKFR rating.
In the controls, I saw that nearly all non TH puzzles were in the 14 451 700.

Revising the updating process, I think now that a unique data base of minimals for TH and non TH will be simpler.

The target is now a fix length format, easier to process

The start of a record (SS) will be

Puzzle; Puzzle Name as {solution grid rank;puzzle bit field}

Code: Select all

................31.98.1.6..2.1..4....871..4......78..6.74..19..5.29...6.....4....;1069081335;00mMf4E1ccfHG0


Or the smaller start with the name only

Code: Select all

1069081335;00mMf4E1ccfHG0


Then 5 fields

SS;25;105;a;b;cccc

[list=]25 number of clues
105 skfr ratingx10
a corner box if a TH pattern has been assumed (or forced later)
b second box if it is an assumed TH with one of the 3 digits in another box
cccc digits having the potential to be in a TH threat
[/list]

a,b,cccc are first given by my TH analysis, but can be forced later to other values

a is
0 if it is an assumed non TH record
1-9 for the corner box of an assumed TH record

b is usually 0, 1-9 if the given for the 3 digits are in 2 boxes

cccc is a 4 bytes field with a leading blank + 3 digits for a TH assumed record
cccc contains first the digits fitting with the test for a potential TH threat.
If the number of digits is <3 ccc is set to ----
4 digits will be as far as I can see for puzzles having a potential for a 4 digits residual pattern.

Later, for a non TH puzzle, this field could contain digits of interest

Some records without the puzzle

Code: Select all

1069081335;00mMf4E1ccfHG0;25;105;0;0;----
1034697749;00m352fEeRm6j1;28;105;3;0; 126

 
847921971;00EaQci19C94Y4;25;110;0;0;2367
204152169;00EqK90c9COKI2;24;105;0;0;2567

I wait for the end of the re rating of the file of minimals, (50% done) , the fully updated file will be released in my google drive as a dy2603 likely end of next week

[champagne]

Sorry for the delay, but I wanted to get a clean file to start the new design.

After hard time trying to update the potential hardest status with the split TH/NoTH, I came back to a unique data base with fields making easier the extraction of puzzles with a specific property.

The first file is done of minimal ED puzzles with several fields

Puzzle,
name as solution grid rank + 81 bits field
Number of clues
skfr ER
3 fields to describe an assumed TH or an assume nEnd digits threat.

I updated the readme file on my google drive in line with this new design.

I uploaded on the drive the dy_ph_2603 file containing only, so far, the file of 14 451 700 puzzles.
The file of expanded will come later this month.

The released file contains
10 486 604 puzzles with a suspicion of TH threat
334 198 puzzles with a suspicion of nEnd pattern. An analysis of these puzzles will come in the relevant thread. The last post of "marek stefanik" shows that the filter for a poten tial nEnd status catches also some exocet patterns!!

[champagne]

I added to the .zip dy_ph_2603 the file phmBexnosub.txt containing the expanded puzzles with no subset for the solution grids with more than 3 minimal.

Are missing in this file the skfr rating and the TH/NOTH analysis, to come in the next days.

The file has 1 111 575 records for a start of 13 139 799 records.

The final status for the current list of minimal is now ok, except for the missing fields.

[champagne]

the file phmBexnosub.txt has been replaced by the final file phex2603db including the rating and the TH/NOTH analysis.
The .zip file is now as expected, with both files sorted on the solution grid rank as first field.