Talk:Monte Carlo methods: Difference between revisions

← Older edit

Talk:Monte Carlo methods (view source)

Revision as of 15:32, 30 May 2018

3,762 bytes added , 6 years ago

→‎Error in all languages

Anonymous user

rosettacode>Oldboy 1948

Revision as of 00:21, 16 November 2016 (view source) Hout (talk \| contribs) (→‎Formulae hidden to most browsers by under-tested cosmetic edits at 18:00, 11 September 2016) ← Older edit		Latest revision as of 15:32, 30 May 2018 (view source) rosettacode>Oldboy 1948 (→‎Error in all languages)
(3 intermediate revisions by 2 users not shown)
Line 44: : Now repaired [[User:Hout\|Hout]] ([[User talk:Hout\|talk]]) 00:20, 16 November 2016 (UTC) ==other REXX (output) examples== Here are some other REXX examples (runs): <pre> scale: 1 234567890123456789012345678901234567890123456789012345678901234567890123 true pi= 3.141592653589793238462643383279502884197169399375105820974944592307816406+ times= 5000000000000 100,000 repetitions: Monte Carlo pi is accurate to 3 places. 200,000 repetitions: Monte Carlo pi is accurate to 4 places. 2,700,000 repetitions: Monte Carlo pi is accurate to 5 places. 2,800,000 repetitions: Monte Carlo pi is accurate to 6 places. 17,700,000 repetitions: Monte Carlo pi is accurate to 8 places. 682,600,000 repetitions: Monte Carlo pi is accurate to 10 places. </pre> <pre> scale: 1 234567890123456789012345678901234567890123456789012345678901234567890123 true pi= 3.141592653589793238462643383279502884197169399375105820974944592307816406+ times= 5000000000000 100,000 repetitions: Monte Carlo pi is accurate to 4 places. 700,000 repetitions: Monte Carlo pi is accurate to 5 places. 6,000,000 repetitions: Monte Carlo pi is accurate to 6 places. 6,100,000 repetitions: Monte Carlo pi is accurate to 7 places. 133,700,000 repetitions: Monte Carlo pi is accurate to 8 places. 362,300,000 repetitions: Monte Carlo pi is accurate to 9 places. 623,300,000 repetitions: Monte Carlo pi is accurate to 10 places. </pre> <pre> 1 2 3 4 5 6 7 scale: 1 234567890123456789012345678901234567890123456789012345678901234567890123 true pi= 3.141592653589793238462643383279502884197169399375105820974944592307816406+ times= 5000000000000 100,000 repetitions: Monte Carlo pi is accurate to 4 places. 12,900,000 repetitions: Monte Carlo pi is accurate to 5 places. 36,200,000 repetitions: Monte Carlo pi is accurate to 6 places. 38,400,000 repetitions: Monte Carlo pi is accurate to 7 places. 69,500,000 repetitions: Monte Carlo pi is accurate to 8 places. 476,200,000 repetitions: Monte Carlo pi is accurate to 10 places. 1,959,100,000 repetitions: Monte Carlo pi is accurate to 11 places. 4,006,300,000 repetitions: Monte Carlo pi is accurate to 12 places. </pre> <pre> scale: 1 234567890123456789012345678901234567890123456789012345678901234567890123 true pi= 3.141592653589793238462643383279502884197169399375105820974944592307816406+ times= 5000000000000 100,000 repetitions: Monte Carlo pi is accurate to 3 places. 900,000 repetitions: Monte Carlo pi is accurate to 4 places. 2,100,000 repetitions: Monte Carlo pi is accurate to 5 places. 4,900,000 repetitions: Monte Carlo pi is accurate to 6 places. 471,400,000 repetitions: Monte Carlo pi is accurate to 7 places. 471,600,000 repetitions: Monte Carlo pi is accurate to 8 places. 491,100,000 repetitions: Monte Carlo pi is accurate to 9 places. </pre> == Error in all languages == Hello! Regarding Monte Carlo methods the results are not very good in most languages. You can even get worse result with more samples. Please if anyone with insight could explain this! Have you tested RDRAND and Monte Carlo methods? I have found some results on MASM32.com. Two links: http://masm32.com/board/index.php?topic=2441.0 http://masm32.com/board/index.php?topic=2432.0 The last is for 64 bit and the mc1.zip is OK. It looks good for Intel Ivy Bridge but not so good for AMD (slower and higher error). Best regards Oldboy 1948