Originally posted by Nans DESMICHELS
Nope, verify by yourself, the pattern end clearly with 14, and it has no correlation with the kind of calculator you use.![]()
He is telling me that I should have better result with a double precision calculator than with a single precision one. He will never suppose think that I've tryied yet, with Simple, Double and infinite precisions.
[edit on 31-7-2004 by Nans DESMICHELS]
; epsilon(10^-1000)
; display(1010)
; acos(-1) 3.141592653589793238462643383279502884197169399375105820974944592*64062862089986280348253421170679821480865132823066470938446095505822317253594081 284811174502841027019385211055596446229489549303819644288109756659334461284756482337867831652712019091456485669234603486104543266482133936072602491412 737245870066063155881748815209209628292540917153643678925903600113305305488204665213841469519415116094330572703657595919530921861173819326117931051185 480744623799627495673518857527248912279381830119491298336733624406566430860213949463952247371907021798609437027705392171762931767523846748184676694051 320005681271452635608277857713427577896091736371787214684409012249534301465495853710507922796892589235420199561121290219608640344181598136297747713099 605187072113499999983729780499510597317328160963185950244594553469083026425223082533446850352619311881710100031378387528865875332083814206171776691473 035982534904287554687311595628638823537875937519577818577805321712268066130019278766111959092164201989
Here I calculated Pi to 1000 decimal places. As you can see, I have no lack of precision
; Pi=acos(-1)
; Pi/2-atan(142857)
0.0000070000069998926663306693490823372643966250467550300016275153447109769632027978000917349908648380493515053459880724627083355712766058481055139495 833747343211781148401610841901839501513714938378133996906295954105516450689045096684454031155862459047721423013172684367039225914289317925365859961239 237790324573489579739239151894020082213338338238569482179083733657189354569494140342550630647305776649173488085440405889493795748406877339035320132220 323988096164879692234443058036418375577188270222971007253736644608395798636379354238346033523136019263191477583424644274233357011005516846749222806464 259116870007250706337280482940601277401176405057217297945365728449891369558132616479980922966916098671443983502690195731739659962883187320899978730473 756299151325754544153057922472973064644574305135670795216582080465069938020247163087523273707352578402359158249235842729724842656974893990886484775133 5894906048522122834484975798298278574237631181981813735909533636692023662371297299816079820211457962915
; Pi/2-atan(14285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714285714 2857142857142857142857142857142857142857142857142857142857142857142857142857142857)
0.0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000000000000070000000000000000000000000000000000000000000000000000000000000000000000000000 000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000069999 999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999 999999999999999999999999999999999999999999999999999999999999999998926666666666666666666666666666666666666666666666666666666666666666666666666666666666 666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666663306666666666 6666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666666665
Clearly the pattern doesn't terminate regularly. Perhaps you can tell me what calculator or software you are using and I can solve your problem?
Also, I did the same experiment with 10,000 decimal places here. You can verify that my calculator got Pi correct by searching "digits of Pi" on google and comparing.
Also, even without numerical computation, amantine's last post has the correct explanation, using atan's Taylor series, why the result will never be regular.



