Re: Strange behavior of NSum on 4.0
- To: mathgroup at smc.vnet.net
- Subject: [mg28080] Re: Strange behavior of NSum on 4.0
- From: Jens-Peer Kuska <kuska at informatik.uni-leipzig.de>
- Date: Fri, 30 Mar 2001 04:12:30 -0500 (EST)
- Organization: Universitaet Leipzig
- References: <99uruh$5vv@smc.vnet.net>
- Sender: owner-wri-mathgroup at wolfram.com
Hi,
a) Plus @@ a
wil compute the sum
b) NSum[] uses the first few terms to approximate a
large sequence of terms by an extrapolation algorithm.
and I expect that 25 terms are the boundary when
NSum[] switch from native summation to the extrapolation
Regards
Jens
Ken Kikuchi wrote:
>
> I encountered a strange behavior with NSum on Mathematica 4.0 on Mac.
> The following is a simplified example about sum of 1 to 25.
>
> In[1]:= a = Range[25]
> Out[1]=
> {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
> 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25}
>
> In[2]:= NSum[a[[i]], {i, 25}]
>
> I expected the result as 325 but received the following strange error
> messages.
>
> Part::pspec :
> Part specification 41/2 is neither an integer nor a list of
> integers.
> ...
> NIntegrate::inum :
> Integrand <<1>>[[41/2]] is not numerical at {1}={41/2}."
>
> However, the same code works fine up to 24.
> In[23]:= NSum[a[[i]], {i, 24}]
> Out[23]= 300.
>
> It works fine if not using Part.
> In[3]:= NSum[i, {i, 25}]
> Out[3]= 325.
>
> But Part works fine with N while manual noted -- N[Sum[...]] calls NSum.
> In[4]:= N[Sum[a[[i]], {i, 25}]]
> Out[4]= 325.
>
> And Sum works too.
> In[5]:= Sum[a[[i]], {i, 25}]
> Out[5]= 325
>
> My version is,
> In[6]:= $Version
> Out[6]= "4.0 for Power Macintosh (July 26, 1999)"
>
> I am not sure it's only my environmental problem or my incorrect usage.
> Could anyone help me?
>
> Many thanks,
> Ken