How I Create a Colab Session That Factors factordb Composites with YAFU

[EdH]

Originally Posted by bsquared

Are colab sessions single threaded? If not it would be helpful to run multithreaded.

There are two threads. I thought I was running two:

    yafuini.write("threads=2\n")

should provide two threads. It used to anyway. I'll check again in a bit.

[bsquared]

Originally Posted by EdH

There are two threads. I thought I was running two:

    yafuini.write("threads=2\n")

should provide two threads. It used to anyway. I'll check again in a bit.

Ah, I missed that line. Thanks.

[Harvey563]

I am getting the following warning in the runtime log:

Jul 26, 2021, 5:43:30 AM WARNING sh: 1: /content/ecm: Permission denied

Is this a problem?

"!chmod -R 777 /content/ecm" doesn't help.

Thanks.

[Harvey563]

I'm not seeing an errors in results, just wondering.

[EdH]

I might be misunderstanding something (probably), but I'm not finding a runtime log or an error message. However, the third candidate in my test run failed and I'm not sure why. That composite is made up of a bunch of p3s and p4s and YAFU factored it in the Colab session when I ran it outside of the script.

sample run of script:

Installing system packages. . . 
Retrieving GMP-ECM. . . 
Compiling GMP-ECM. . .
Finished installing GMP-ECM. . . 
Retrieving YAFU. . . 
Compiling YAFU. . . 
Finished compiling YAFU. . . 
Starting the factoring of 3 composites. . .  

Composite 1: 8869475002717536050312782604253426183326192888885941092918900891116352039782449522669 <85> 
Factors (6:59): 3211610951880144183669785219693807857*2761690359015982301564877430778307467378692976317  

Composite 2: 1128164293201891394597728447518614695546967713433260789629082173322486406149926324143 <85> 
Factors (6:02): 29751725740768035469240132235509074973407496022009533350217*37919289221465108716321079  

Composite 3: 3895909508598394792966931237542439743139398349882487979187550567767896277846662661167 <85> Factors (0:00):   

Completed all 3 composites!

factor log of third candidate run in isolation:

fac: factoring 3895909508598394792966931237542439743139398349882487979187550567767896277846662661167 
fac: using pretesting plan: normal 
fac: no tune info: using qs/gnfs crossover of 95 digits 
fac: no tune info: using qs/snfs crossover of 75 digits 
div: primes less than 10000 
Total factoring time = 0.0034 seconds 
 
 
***factors found*** 
 
P3 = 911 
P3 = 941 
P3 = 953 
P3 = 967 
P3 = 971 
P3 = 977 
P3 = 983 
P3 = 991 
P4 = 1009 
P4 = 1019 
P4 = 1031 
P4 = 1033 
P4 = 1039 
P4 = 1049 
P4 = 1061 
P4 = 1063 
P4 = 1087 
P4 = 1091 
P4 = 1093 
P4 = 1097 
P4 = 1103 
P4 = 1109 
P4 = 1117 
P4 = 1129 
P4 = 1151 
P4 = 1153 
P4 = 1163 
P4 = 1171 
1

[chris2be8]

Originally Posted by Harvey563

I am getting the following warning in the runtime log:

Jul 26, 2021, 5:43:30 AM WARNING sh: 1: /content/ecm: Permission denied

Is this a problem?

"!chmod -R 777 /content/ecm" doesn't help.

Thanks.

Run "!ls -l /content/ecm" to see what /content/ecm really is (I hope that's the right syntax). HTH

[Harvey563]

!ls -l / content/ecm results:

total 18512 -rw-r--r-- 1 root root 21167 Jul 27 19:23 acinclude.m4 -rw-r--r-- 1 root root 43125 Jul 27 19:23 aclocal.m4 -rw-r--r-- 1 root root 36281 Jul 27 19:23 addlaws.c -rw-r--r-- 1 root root 2672 Jul 27 19:23 addlaws.h -rwxr-xr-x 1 root root 456488 Jul 27 19:23 aprcl drwxr-xr-x 3 root root 4096 Jul 27 19:23 aprtcle drwxr-xr-x 2 root root 4096 Jul 27 19:23 arm drwxr-xr-x 2 root root 4096 Jul 27 19:23 athlon -rw-r--r-- 1 root root 1758 Jul 27 19:23 AUTHORS drwxr-xr-x 2 root root 4096 Jul 27 19:23 autom4te.cache -rw-r--r-- 1 root root 2149 Jul 27 19:23 auxarith.c -rw-r--r-- 1 root root 8124 Jul 27 19:23 auxi.c -rw-r--r-- 1 root root 7390 Jul 27 19:23 auxlib.c ...

It appears to be source code.

"View runtime log" is the final option under runtime menu in Colaboratory.

[EdH]

Originally Posted by Harvey563

I am getting the following warning in the runtime log:

Jul 26, 2021, 5:43:30 AM WARNING sh: 1: /content/ecm: Permission denied

Is this a problem?

"!chmod -R 777 /content/ecm" doesn't help.

Thanks.

OK, I think I have it fixed. Thanks for pointing it out to me. (I guess I'm going to have to look at all my other Colab threads that use ECM, as well.)

To manually fix, change the line:

     print(line.rstrip().replace('ecm_path=../gmp-ecm/bin/ecm', 'ecm_path=/content/ecm'))

to

    print(line.rstrip().replace('ecm_path=../gmp-ecm/bin/ecm', 'ecm_path=/usr/local/bin/ecm'))

Remember to keep the four leading spaces. Also remember that if you've already run the session without the change, you will need to terminate and reconnect the session to clear all the original work.

[Harvey563]

This fixed it. Thanks.

[Happy5214]

The script as written does not report factors found via TF (i.e. those less than 10^5). This code fixes that issue (replace lines 85-89 with this code):

  prp_string = (", prp")
  tf_string = (", div: found prime factor")
  fcheck = 0
  factors = ""
  for line in file:
    found = line.rfind(prp_string)
    if found < 0:
      found = line.rfind(tf_string)

[EdH]

Changes made. Thanks Alexander