In order to use the Demography Lab, you must have an account. You probably have one already, but if not please visit http://lab.demog.berkeley.edu/LabWiki/index.php/Main/GettingStarted AND THEN download and fill out the http://lab.demog.berkeley.edu/Docs/statementofcompliance.pdf.
You can find lots of good information at http://lab.demog.berkeley.edu/LabWiki. Check it out in your copious free time.
The textbook for this course is site licensed by UCB for students. In order to access the book, your computer must either - have an ipaddress associated with UC Berkeley - the Library’s ezproxy enabled https://guides.lib.berkeley.edu/ezproxy/browser-bookmarklet
If you are connecting to the internet via AirBears or through a workstation in the basement of 2232, then your ipaddress will get you in. If you are elsewhere in the universe, you’ll need the ezproxy.
We use noMachine to get a complete Demography Lab desktop when we’d prefer not to be in the attic or basement of 2232 Piedmont. noMachine is very clever software – the remote desktop that it provides is fast and unbreakable. Even if the internet breaks for a week, your noMachine session remains active and you can just reconnect to it and pick up where you left off.
We do not need to use noMachine today because R/Rstudio is the only Demography Lab application that we’ll use. For that we can simply browse to https://rstudio.demog.berkeley.edu But for other stuff – like Stata or Python or SAS or C or Perl or Socsim or anything other than R – noMachine is the best thing ever.
For instructions on how to setup and use noMachine see
http://lab.demog.berkeley.edu/LabWiki/index.php/Main/NoMachine
You’ll want to set this up after class so you can be ready for the rest of graduate school.
The Demography Lab servers all run Linux (a popular variant of the Unix operating system). Unix is also the underlying OS of many popular Apple products e.g MacOS. Thus many of the cool tricks that you can do on Linux machine, can also be done on Macs.
Although Linux machines, like all modern OS’s support graphic user interfaces (desktops), Linux, in it’s purest form runs at the command line. This is good thing for scientists because it makes us stronger and much more efficient. For a gentle introduction to Linux at the Demography Lab, please take a look at The 12 most important Linux commands http://lab.demog.berkeley.edu/pmWiki/LabWiki/uploads/Main/12important.pdf
Since we will not be using anything other than R/Rstudio this week, Linux commands will not be crucial right away, but check out the above document after class so you know what lies ahead.
Rstudio is the application that we will live in this term. Before we dive in, Let’s note a couple of things:
Rstudio is separate from R. Rstudio is an “IDE” (integrated development environment) in other words, it’s a program that helps you program. R, on the other hand, is essentially a computer language - it executes code and that’s it. So when you are running Rstudio, you are generally running TWO separate processes: R and Rstudio. While Rstudio understands this, R does not.
rstudio.demog.berkeley.edu runs in the cloud it therefore does not have direct access to your local filesystem instead, it has access to your Demography Lab account and thus to the entire Demography Lab filesystem. This means that if you want to read files they must be on the Demography Lab filesystem and any files you write will be on the Demography Lab filesystem.
rstudio.demog.berkeley.edu runs on our server, keyfitz. keyfitz is a respectable machine with 250GB of RAM and 28 Cores. That’s more than most demographers need individually, but collectively – more is always better. Nothing you do in Demography 213 is likely to consume enough of keyfitz’s resources for anyone else to notice, however, other scientists who share the system might on occasion slow things down with their tireless explorations of flat likelihood surfaces or simulation models of the universe. If you notice this sort of thing, let me know.
rstudio.demog.berkeley.edu comes with a default option that is bad for your soul. To fix this you should do the following (only once)
Goto Tools -> Global Options
Under the "General" Tab:
uncheck the box : Restore RData into workspace at startup
and change "Save workspace to RData on exit" to "NEVER"Make this change or suffer the consequence – irreproducible results.