Re: How to increase Working Precision?

On Mar 5, 4:30 pm, "Nasser Abbasi" <n...@xxxxxxxxx> wrote:
"rjf" <fate...@xxxxxxxxx> wrote in message

news:c742b996-12f0-471a-b1d6-c2a53cd57aa9@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

On Mar 5, 12:35 pm, "Nasser Abbasi" <n...@xxxxxxxxx> wrote:
"Paul J Salmon" <pauljsal...@xxxxxxxxx> wrote in
messagenews:8b028ca7-4e58-4167-95ea-ad06d5428c4f@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

2.0 is low precision. try N[Pi+2,1000]. You could try reading the
manual.

Using N[] that way seems to be the same as using SetPrecision:

In[43]:= r = N[Pi + 2, 1000];
Precision[r]
Out[44]= 1000.

In[45]:= r = SetPrecision[Pi + 2, 1000];
Precision[r]
Out[46]= 1000.

What exactly is the difference you show by using N[] vs. SetPrecision[] as I
said?

Nasser

Try

In[63]:= Zeta[18, 3]

Out[63]= -(262145/262144) + (43867 \[Pi]^18)/38979295480125

In[64]:= SetPrecision[%, 25]

Out[64]= 2.59599937483986*10^-9

In[65]:= N[%%, 25]

Out[65]= 2.595999374839856461644622*10^-9

you could see that N internally increased settings of SetPrecision
so that the result was correct to 25 requested digits:

In[67]:= SetPrecision[-(262145/262144) + (43867 \[Pi]^18)/
38979295480125, 35]

Out[67]= 2.595999374839856461644622*10^-9

Hope this helps,
Oleksandr Pavlyk
Wolfram Research
.

Loading