2022 “small” 15e post processing reservations and results

[chris2be8]

Originally Posted by kruoli

The msieve.log is over 10 MB, unfortunately. Should I include it by stripping it from some information?

What makes it so big? If it's lots of "error -nn reading relation nnnn" messages then the following script will summarise them:

#!/usr/bin/perl

# This summarises msieve.log ready to post.

use warnings;
use strict;
use LWP::Simple;

my $logfile=$ARGV[0]; # The log to read
my $numerrors=0; # Number of errors
my %errors = (); # How many errors of each type

if ( ! -f $logfile) {
  warn "*********** $logfile does not exist ***********\n";
  exit;
}

warn "*********** Processing $logfile *************\n";
open(INFO,$logfile) or die "Can't read $logfile $!";
while(<INFO>) {
  if ($_ =~ /error -(\d+) reading relation /) {
    # found a url to post
    $numerrors++;
    $errors{$1}++;
    next;
  }
  if ($numerrors) {
    print "[ . . . skipped $numerrors error -n reading relation nnn messages . . . ]\n";
    $numerrors = 0;
    my $err;
    foreach $err (sort by_number keys %errors) {
      print "[ $errors{$err} error -$err ]\n"
    }
    %errors = ();
  }
  print;
}
close(INFO) or die "Error closing $logfile $!";

sub by_number { $a <=> $b }

If it's not that then just posting the first and last few messages to skip whatever mesage is repeated would be reasonable.

[kruoli]

Yes, I had an absurd amount of relations that could not been read. But I just saw they were largely from previous attempts when they haven't been enough relations for a matrix. So I stripped it down drastically to only the mattering invocations. I should have deleted the logs of the unsuccessful attempts.

[swellman]

11p8_254 is factored.

p76 factor: 2963897012281681678709613115834105724195821421454260604345654902584241693501
p137 factor: 42096691579079820095319625287723535446122231598963322246523234962616534155708894015439747745944923458282516446628478540382633789265057841

https://pastebin.com/FtXevqTJ

[bur]

Forgot to mention: Reserve C174_1992_1695.

[richs]

C167_230916_6623 has built a matrix. Results 9 September AM.

[richs]

p55 factor: 1839740722401033700417540068375751085542761086256704669
p113 factor: 37223169195598119704454339139486327729543836485915375288944005642702463758420855534496238295602229908929790031581

Approximately 25 hours on 6 threads of a Core i7-10750H with 16 GB memory for a 4.92M matrix at TD=130.

Log attached and at https://pastebin.com/URFX7QkH

Factors added to factorDB.

[kruoli]

4_430m has built a 12.7 M matrix at TD=140 and can be move to post-processing.

[frmky]

34210243_31m1 is done. LA took about 30 minutes.

[bur]

matrix is 8306661 x 8306886 (3805.9 MB) with weight 1017736557 (122.52/col)
C174_1992_1695 sufficient relations, can stop sieving.

It had a uniques rate of 80%, which I found quite high. Any idea why that happened?

[kruoli]

Your Q range is really small, that helps with uniques.

[bur]

So in a way, the higher the raw yield is, the higher the uniques:duplicates ratio will be?