Hi all,
Sorry to bother everyone after class is over - I was wondering if anyone
knows how/has any code to create 95% confidence intervals for a lowess line
in R? I've seen these done in articles, and am wondering how to do it..Feel
free to reply to me directly rather than the whole class to avoid clogging
up people's inboxes...
Cheers,
Mark
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
Gov2001,
I hope you all are doing well as you prepare for your other exams
and/or finish the replication paper.
I wanted to give you guys an update on the code library. First let me
say that you guys did a great job, but editing it became considerably
more time consuming than we anticipated. While I wouldn't describe it
as "done" its at a stage where we believe it can be released (hence we
are calling it version .5 to indicate its incompleteness, we will put
out an announcement through the list when we release a new version).
Basically we hit about the 25+ hour mark of working on it and decided
that was enough for the time being. The final thing clocks in at 45
topics (including 10 that weren't on the list you guys signed up for),
14 datasets, 7 script files and over 5500 lines. Cool!
It is posted on the Section area of the website. We will continue to
develop and expand it. To this end, two requests:
1) If you have some cool code that you used in your replication paper
that you'd like to share with others, send it to me (not the list) and
I will include it in the next version.
2) If you have any suggestions for topics not covered similarly send
me an email.
If you happen to review the code you wrote, you may find that it is
considerably different. In some cases pieces of coded needed to be
corrected or clarified (in a few rare cases completely rewritten).
We think that what remains is fairly complete and accurate but do let
us know if you find any errors.
Thanks again for participating in this project. We hope it will be a
useful resource to you as well generations of students to come.
Brandon and Molly
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
--
gov2001 mailing list served by Harvard-MIT Data Center
List Address: gov2001(a)lists.gking.harvard.edu
Subscribe/Unsubscribe: http://lists.gking.harvard.edu/?info=gov2001
This looks like a terrific project on statistics and privacy (how to make
available statistical results of confidential data sets so that no research
subject can be identified). please contact Salil if you are.
Gary
--
*Gary King* - Albert J. Weatherhead III University Professor - Director,
IQSS - Harvard University
GKing.Harvard.edu <http://gking.harvard.edu/> - King(a)Harvard.edu -
@kinggary<http://twitter.com/kinggary>- 617-500-7570 - Asst 495-9271 -
Fax 812-8581
---------- Forwarded message ----------
From: Salil P. Vadhan <salil(a)seas.harvard.edu>
Date: Sat, Apr 30, 2011 at 10:39 PM
Subject: RE: looking for a student
To: Gary King <king(a)harvard.edu>
Oops – yes, it is paid. I added a sentence to that effect. Thanks!
Internship in Privacy-Preserving Statistical Estimation
----------------------------------------------------------------
We are looking for an intern in Summer 2011 to implement and test some new
algorithms for incorporating privacy into statistical estimation.
Specifically, the intern will implement some privacy-preserving wrappers
around Zelig methods, and run experiments to measure their effect on the
quality of results obtained. These wrappers will be based on recent work
in “differential privacy,” which is a framework for ensuring that no
individual’s data has a noticeable influence on the result of a statistical
analysis.
Relevant skills include comfort with programming in R and background in
statistical estimation. The intern will receive a stipend, at a rate to be
determined based on level of experience.
This internship is part of the “Privacy for Social Science Research”
project, which is a multidisciplinary collaboration between the Center for
Research on Computation and Society, the Institute for Quantitative Social
Science, and the Berkman Center for Internet and Society. (See
http://crcs.seas.harvard.edu/research-areas/)
Interested candidates should email the following to Salil Vadhan (
salil(a)seas.harvard.edu) by noon on May 9:
- Resume
- Transcript
- Names and email addresses of 2-3 references
- The earliest date you are available to start work on the project
- A few sentences about your interest in and qualifications for
this specific project
*From:* Gary King [mailto:king@harvard.edu]
*Sent:* Saturday, April 30, 2011 10:33 PM
*To:* Salil P. Vadhan
*Cc:* Edo Airoldi; Jonathan Ullman
*Subject:* Re: looking for a student
Salil, is the internship paid?
Gary
--
*Gary King* - Albert J. Weatherhead III University Professor - Director,
IQSS - Harvard University
GKing.Harvard.edu <http://gking.harvard.edu/> - King(a)Harvard.edu -
@kinggary<http://twitter.com/kinggary>-
617-500-7570 - Asst 495-9271 - Fax 812-8581
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
