See post #527.
]]>You didn't tell me how we actually get the numerical answer here.
]]>I set a[n] to be the array of the series terms. Then I do N[acc[number_of_terms],number_of_digits.
]]>accelerate[n_]:=Module[{d,b,c,s},
d=(3+Sqrt[8])^n;
d=(d+1/d)/2;
b=-1;
c=-d;
s=0;
Table[c=b-c;s=s+c*a[k];b=(k+n)(k-n) b/((k+1/2)(k+1)),{k,0,n-1}];
s/d]
Remember how to use it?
]]>Can you post the accelerator Borwein found?
I am sure I showed it to you already. I know we discussed it. It is also useless for this sequence. I used Romberg and it worked well.
Want the code for Borwien's?
]]>Can you post the accelerator Borwein found?
]]>Look at the latest one discovered by the Borweins. I am not saying you should change what you wanted to do, just that this is an important and open field. I did not think of this when you asked me for a project.
But we are getting far off track of the point of this thread.
]]>