This is a repository for the latest version of the ARE212 section notes. Each section has its own directory that contains at least three core files:
An org-mode document
.orgthat compiles to the
.Rfiles. In fact, the org-mode document is the code; and you can dynamically update all downstream documents from within Emacs' org-mode. You do not have to interact with the org-mode document directly if you are not using Emacs, but rather just with the R code or PDF write-up. If you'd like to get set up with Emacs (highly recommended) then please see the next section of this README.
A PDF of the section notes, which effectively documents the code. If you are only interested in following along, rather than running the code yourself, just browse to the
sec-01.pdf) and click "view raw". The PDF will begin downloading immediately.
An R script that compiles all of the code within the PDF. Note that there is no documentation within the code. Instead, the code is documented from the PDF description.
If there are supporting images or TeX fragments for the write-ups,
there will be a subdirectory called
inserts/ within the section
The sections are organized as follows:
section-01 Preliminaries and setup
section-02 Matrix operations in R
section-03 OLS regressions from first principles
section-04 Goodness of fit
section-05 Hypothesis testing
section-06 Returns to education, empirical example
section-07 Efficiency of GLS and
section-08 Instrumental Variables
section-09 Testing for heteroskedasticity
section-10 Feasible generalized least squares
section-11 Serial correlation
section-12 Instrumental variables
section-13 Spatial analysis in
section-14 Web scraping
Help me write this!
This project can and should be treated like any other open source, collaborative coding project. If you are interested in helping me make this project better, fork the repo, edit the screwy files, and send a pull request. I will review and merge the changes -- until someone else takes over!
Org mode notes
If you are running Emacs, then
you have access to org-mode, an open source
solution for interactive coding and reproducible research. The code,
documentation, and results are all bundled into the same file. The
#+RESULTS output is automatically generated from the immediately
preceding code block.
A few things to note. When you try to compile the
.org files to
a PDF document, you may have to compile it twice or reload the buffer
C-u M-x org-reload. To tangle the code within the org-mode
document to an
.R script, you can use the key binding
C-c C-v t.
You can highlight code by using the
minted package in LaTeX. For
this, from the command line, make sure that you invoke
-shell-escape flag. For example,
cd ~/Dropbox/github/danhamer/ARE212/section-04 pdflatex -shell-escape sec-04
Ensure that you have the proper
minted.sty file by downloading the
zipfile, installing it, and then ensuring that LaTeX knows where
unzip minted.zip cd minted/ latex minted.ins sudo texhash
Finally, you will have to add the following to your
file, and make sure it doesn't conflict with anything else in there:
(require 'org-latex) (setq org-export-latex-listings 'minted) (add-to-list 'org-export-latex-packages-alist '("" "minted"))
Spatial analysis in R
This is of personal interest. R is ideal for econometric analysis; but it also has some very convenient facilities for interacting with relational databases that support spatial data analysis. A notable example is the open source project cartodb-r. An example of the type of spatial data analysis that can be done from within R, riding on CartoDB is shown below.
The orange and blue points are households in New Delhi; the orange indicates a relatively healthy household, the blue indicates a household where at least one member has recently experienced diarrhea. The green points are sewage and garbage facilities. This is a sort of modern-day cholera map.