Hey guys,
I keep getting an error message when I include "hessian=TRUE" in my optim
command...it says object "hessian" not found. Does anyone else have this
problem/know how to fix it...It seems to work fine without the hessian
term...
Best,
Abby
Our group would like some advice on selecting a paper for replication. Based
on the guidelines it seems that we would be best off selecting a paper the
misuses relatively complex methods, yet has been published in a major
journal. Is this similar to what other people have been looking
for/selected in previous classes? Also, how similar should our methods be to
the paper we are replicating?
Ben Snyder
Anybody know how to use the optim() command on a function that sends more than one argument into it? For example, I created a function:
Sample <- function(a,b,c) {
... }
I'd like to optimize parameter "a" in the function sample, while entering data for parameters "b" and "c" into the optimization. Any suggestions? I can't quite get it to work.
Thanks in advanced.
Anil Doshi
Doctoral Student | Technology and Operations Management
Harvard Business School
302 Wyss Hall
Boston, MA 02163
tel 646-244-5396
email adoshi at hbs.edu<mailto:adoshi at hbs.edu>
Hi everyone,
To make up for missing office hours on Monday because of the holiday, I'm
going to hold office hours Tuesday, 10-12 for anybody who is interested.
They will be in the usual location. Also, there will be no lecture
tomorrow.
Best, Iain
In problem 3 of the current problem set, I think there is a typo in the
definition of f(x,y). Going on the R-code, a factor of 1/2 has been missed
from the power of the exponential.
i.e. f(x,y) = e^[*-0.5**((x-2)^2+(y-1)^2)]
Also in the following text, I think it should say with equal variance
sigma_x=sigma_y=1 rather than sigma.
In notational terms, in problem 2.3 I think it should read L(pi|y) rather
than L(beta|y)?
Cheerio, Nicola
Hi,
for question 2.3: when we are looking for the bearded man's approximation we need to use the likelihood function as our f_x, so do we need to find the first and second derivative of the binomial likelihood function conditional on our 5 data points analytically and evaluate it at the MLE we found?
Or should we use the log-likelihood function and its first and second derivatives?
Thanks a lot,
Chiara
Hi everybody,
I have a question on 2.3 and 2.4 in Assignment 3.
In 2.3, I did a quadratic approximation using Taylor expansion of the "likelihood" function, not "log-likelihood" function.
In 2.4, I was asked about a quadratic approximation to the "log-likelihood" even though we have never explored it in previous questions.
Do you have any idea on that, or how can I solve it?
Have a nice weekend!!
Akihiro in Longwood..
>>> "Lin, Eric" 02/11/10 1:31 PM >>>
I?m trying to typeset a long equation in an eqnarray environment. I?m
trying to figure out how to wrap the right side of the equation such that it
wraps on on the right side of the first equals sign:
Y= a + b + c
+d +e +
Any ideas on how to do this within an eqnarray environment?
Thanks!
EXL
_______________________________________________
gov2001-l mailing list
gov2001-l at lists.fas.harvard.eduhttp://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
I?m trying to typeset a long equation in an eqnarray environment. I?m
trying to figure out how to wrap the right side of the equation such that it
wraps on on the right side of the first equals sign:
Y= a + b + c
+d +e +
Any ideas on how to do this within an eqnarray environment?
Thanks!
EXL
Hi,
I have minor problem, the margins are quite large. I have tried to use
usepackage{fullpage}
but it can not find fulpage.sys file.
Does anybody know how to adjust margins?
thanks..
Burak Eskici
Doctoral Student, Department of Sociology
Harvard University
563 William James Hall
33 Kirkland Street
Cambridge, MA 02138
eskici at fas.harvard.edu
