Date: Sun, 11 Jul 2010 17:31:37 -0400
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Arthur Tabachneck <art297@NETSCAPE.NET>
Subject: Re: Need to find a value that can make the equation's value = 0
like Solver in Excel
Content-Type: text/plain; charset="utf-8"; format=flowed
I'm cc'ing this to SAS-L as it streches beyond my expertise and think
others might have a better solution.
However, that said, can't you define a close-enough criterion and use
it, in combination with a fuzz function, to stop the process when it
reaches an acceptable solution? Similarly, you could only allow a
maximum number of iterations.
Otherwise, yes, the risk of not being able to resolve such a search can
definitely be problematic.
From: Ching-Hung Chang <email@example.com>
To: Arthur Tabachneck <firstname.lastname@example.org>
Sent: Sun, Jul 11, 2010 5:24 pm
Subject: Re: Need to find a value that can make the equation's value =
0 like Solver in Excel
Thanks for your response.Â But I have one concern about using the do
while loop as the Ri I am looking for is around 0 to 1.Â And the value
might have to go to something like 0.0012389654788Â in order to solve
the equation.Â So, if I use the do loop to do it, how can I guarantee I
find the right value I want, and it might also take really longer than
I expect when using do loop?Â Thanks.
On Sun, Jul 11, 2010 at 1:32 PM, Arthur Tabachneck
If you only need to solve for one value, couldn't you just a binary-
tree-type-search using a do while loop?
On Jul 11, 1:34Â pm, CC <chchanghe...@gmail.com> wrote:
> On Jul 10, 8:04Â pm, Patrick <patrick.mat...@gmx.ch> wrote:
> > "...need to find Ri for each firm/year."
> > If it's firm/year why then isn't it Ri,t?
> > It would be helpful for someone like me if you also could explain
> > the different literals mean (M, B, ROE,...). I assume ROE is Return
> > Investment, but what's the rest?
> > Was this formula given to you or did you develop it based on a
> > problem? My unqualified impression is that this is more a question
> > getting the maths right first.
> Hi Patrick,
> The formula is a simplified version but I am sure this will not affect
> the result as my main problem is to find a function to deal with it (I
> am pretty sure even it is a simplified version, we can still get an
> answer if we know which function we can use.). Â And, of course, I will
> use the complete version of formula to solve this question later. Â The
> M and B are the market value and book value, respectively. Â I did not
> want to mention it since I will not like to confuse people and want to
> simply focus on the question itself. Â And, yes, it should be Ri,t
> instead of Ri. Â So, your observation is correct. Â Thanks for that.
> But, if we can find one Ri then the rest is easy
> relatively. Â Could you give me some suggestions on how to solve it?