Thanks! Found it.
Sincerely,
Olena Ageyeva
Date: Mon, 2 Mar 2009 22:17:36 -0500
From: plam at
fas.harvard.edu
To: gov2001-l at
lists.fas.harvard.edu
Subject: Re: [gov2001-l] Reparameterization
Hi Olena,
Please view the section video from this past week or check out my section notes posted on
the class webpage. They should clarify things a bit more.
On Mon, Mar 2, 2009 at 10:11 PM, Olena Ageyeva <eageeva at hotmail.com> wrote:
Well, I did reparameterization and now optim() shows different result
for pi_hat. Before I was getting exactly what I expected for that
parameter.
What to do with that? Any clue or where to read about this?
Sincerely,
Olena Ageyeva
From: rolai at
hbs.edu
To: gov2001-l at
lists.fas.harvard.edu
Date: Mon, 2 Mar 2009 21:47:06 -0500
Subject: Re: [gov2001-l] R warnings
check out the pnorm function in R
?pnorm
From: gov2001-l-bounces at
lists.fas.harvard.edu [gov2001-l-bounces at
lists.fas.harvard.edu] On Behalf Of Olena Ageyeva [eageeva at
hotmail.com]
Sent: Monday, March 02, 2009 9:43 PM
To: gov2001-l at
lists.fas.harvard.edu
Subject: Re: [gov2001-l] R warnings
Any clue how to do that? As I understand, I have to constrain the parameter to (0,1)?
Right?
I've never done it before, so it's very new for me.
Sincerely,
Olena Ageyeva
From: rolai at
hbs.edu
To: gov2001-l at
lists.fas.harvard.edu
Date: Mon, 2 Mar 2009 21:30:08 -0500
Subject: Re: [gov2001-l] R warnings
I believe this error occurs when p = 1 (since log(1-p) = -Inf)
make sure you reparamaterize your par!
From: gov2001-l-bounces at
lists.fas.harvard.edu [gov2001-l-bounces at
lists.fas.harvard.edu] On Behalf Of Olena Ageyeva [eageeva at
hotmail.com]
Sent: Monday, March 02, 2009 9:23 PM
To: gov2001-l at
lists.fas.harvard.edu
Subject: [gov2001-l] R warnings
Does anyone know what these warnings
mean and how to fix this?
optim(par, fn=ll.binomial, method="BFGS",
control=list(fnscale=-1), y=y, N=10)$par
[1] 0.7759998
Warning messages:
1: In log(1 - p) : NaNs produced
2: In log(1 - p) : NaNs produced
3: In log(1 - p) : NaNs produced
4: In log(1 - p) : NaNs produced
5: In log(1 - p) : NaNs produced
6: In log(1 - p) : NaNs produced
7: In log(1 - p) : NaNs produced
8: In log(1 - p) : NaNs produced
9: In log(1 - p) : NaNs produced
Thanks!
Sincerely,
Olena Ageyeva
Windows Live?: Life without walls.
Check it out.
Express your personality in color! Preview and select themes for Hotmail?.
See how.
Windows Live?: Life without walls. Check it out.
_______________________________________________
gov2001-l mailing list
gov2001-l at
lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
--
Patrick Lam
Department of Government and Institute for Quantitative Social Science, Harvard
University
http://www.people.fas.harvard.edu/~plam
_________________________________________________________________
Windows Live? Contacts: Organize your contact list.
http://windowslive.com/connect/post/marcusatmicrosoft.spaces.live.com-Blog-…