Hi All,
We've uploaded Problem Set 4, Solution Set 3, today's section handout, and
the R code for today's section to the course webpage.
Following tonight's sections, for those curious to learn more about how
optimization works, the following website lets you experiment with the
BFGS algorithm and can be quite instructive:
http://www.cse.uiuc.edu/eot/modules/optimization/BFGS/
For a problem with multiple modes, try a function such as:
sin(x)+sin(y)
Best,
Dan
----
Ph.D. Student
Department of Government
Harvard University
Tutor, Currier House
dhopkins(a)fas.harvard.edu
http://www.danhopkins.org
Hello,
I was wondering, in problem 2b,c can I safely assume integer values for
my gamma function?
--
Serban Tanasa
Email: tanasa(a)fas.harvard.edu
Cell: 617-833-7847
Address: 513 Cabot Mail Center, Cambridge, MA 02138
Greetings,
I have not been able to get the perspective graph to work. I have
two variables as an output for the multivariate normal distribution.
I don't know what the third variable should be. Should I have more
outputs for the multivariate normal?
Sheldon
Hi all,
Could someone please explain how to read a contour plot as a means of
determining whether there is a relationship between the variables used to
construct it? I understand contour plots in topography, but I don't quite
see the logic as we're using it here. The only thing I can figure is that a
series of concentric rings that are more or less a circle suggest no
relationship (i.e. sigma_{1,2}=0), while a contour plot whose rings form
more of a football shape or cigar shape indicates a stronger relationship.
Is this heading in the right direction? The Cleveland text wasn't of much
help.
Thanks,
Jacqueline
--
Jacqueline Chattopadhyay
Ph.D Student, Government & Social Policy
Harvard University
jchattop(a)fas.harvard.edu
jchattop(a)gmail.com
714.313.7022
the amsmath package is very good with all sorts tricks. See my web page under classes.
Gary
-----Original Message-----
From: Ian Brett Yohai <yohai(a)fas.harvard.edu>
Date: Wednesday, Mar 8, 2006 9:20 pm
Subject: Re: [gov2001-l] LaTeX wrapping equations
Hi Stephen,
Look at the \lefteqn syntax. You might also look at the amsmath package which has more options for this sort of thing.
Best,
Ian
On Wed, 8 Mar 2006, Stephen Purpura wrote:
> How do you convince LaTeX to wrap an equation when it is in a
\begin{eqnarray*) construct?
> My ln() equations run over a single line.
>
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
When imagining the multivariate normal distribution, the sigma matrix
does not seem to be positive definite. This keeps giving me an error
when using the mvrnorm function. Is this an error on the homework I
missed, or am I in left field somewhere.
Sheldon
Hi Ian and Dan,
A quick question about 2a.
Do we have to do all the math involving the derivation of the negative
binomial? Or this is just about showing for the case of that
distribution that we know what the steps of creating our own
distribution are, without actually solving the ugly integral.
Juan
Hi List,
I've been working on problem 3, and I have a handle on how to get the
Inverse CDF from the lecture notes. However, I'm a little lost on the
integration of the Weibull. Should we use the grid method, or go with
traditional integration? I was just wondering, since the examples for the
grid method we have are for joint densities, and the Weibull doesn't seem
to be one (correct me if I'm wrong!)
Thanks,
Emily
Hey all,
Can someone please remind me how to replace all the values within a vector
of numbers that meet a certain criterion with another value? For example, if
we have a vector [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] and want to repalce all of
the values in this vector that are less than or equal to 4 with the number
12, what R code would do this? I've tried to find help in the Fox book and
online and appear incapable of locating it.
Thanks,
Jacqueline
--
Jacqueline Chattopadhyay
Ph.D Student, Government & Social Policy
Harvard University
jchattop(a)fas.harvard.edu
jchattop(a)gmail.com
714.313.7022
I find that my Xemacs R console is filling up with thousands of lines
of garbage (my fault for printing out too much stuff on the screen).
I'd like to be able to clear my console of text so I don't have to
scroll through it.
I remember seeing Ryan Moore do this with couple key strokes last
semester. R help doesn't seem to have much.