Re: [Libmesh-users] Setting a target eigenvalue rather than selecting the position of spectrum

[John P. <jwp...@gm...>]

On Wed, Jun 29, 2011 at 1:56 PM, Omar Al-Abbasi <oma...@gm...> wrote:
> Thanks for the suggestions.
> John are you suggesting that within my Libmesh code and prior to specifying
> the position of the spectrum I shall write ... (-st_shift .75) and then I
> specify the spectrum to LARGEST_MAGNITUDE ??

You specify LARGEST_MAGNITUDE in code, -st_type/shift are *command
line* arguments so when you run your code, you pass them there:

./a.out -st_type sinvert -st_shift .75

assuming your executable was named a.out.  If you want to perform
multiple solves with different shifts during a single run, then you'll
need to do it programmatically, consult the SLEPc docs for more info.

-- 
John