Comments for CityState

Comment on Major improvements to acs.R: sneak peak at version 1.0 by Ezra Glenn

Ezra Glenn — Wed, 10 Apr 2013 18:24:13 +0000

John:

Thanks for your comment, and no worries — I’m glad to hear that you
think the package might be useful. The timeline is now — that
package is already available, and can do a lot already — but it is
also still in development, with new features being add. If you are
familiar with R as a programming environment already, you are good to
go — take a look at the user guide posted to the page you commented
on and you can get going.

As for what’s available, that depends some on the census: the tool
uses the Census API (< http://www.census.gov/developers/>), and so far,
they’ve only made available the 5 year data from 2006-2010 and
2007-2011, and only for certain geographies. As they add more sets,
the package will work for them as well, so my understanding is that it
is worth starting with, and hopefully will not become a dead-end.

Also, to address some of the shortcomings in available data, I’m also
adding in a tool to allow users to more easily download data from the
American factfinder site. They’ll still need to actually click the
download button in a browser, but the R package will do the search and
generate a url for them to do this.

–Ezra

Comment on acs.R: a worked example using blockgroup-level data by Ezra Glenn

Ezra Glenn — Wed, 10 Apr 2013 18:23:16 +0000

Mara:

Thanks for your comment (and glad to see it being used here in my own
backyard!). I haven’t tested this extensively — it really depends
more on the Census API than one the package per se — but from what I
can tell the leading zeroes are optional. When I tried your Chinatown
example, it worked either way.

–Ezra

Comment on acs.R: a worked example using blockgroup-level data by Mara

Mara — Wed, 10 Apr 2013 15:10:26 +0000

I'm a little bit confused about the number of digits to use for a given census tract. To make a custom geography (combining two census tracts) I followed your example from 'Working With ACS R'(p.7) in which you specify the state, county and tracts using 5-digit FIPS. But in order to retrieve the tracts individually I added a leading zero. Are 'leading' zeros optional?

> tract.1=geo.make(state="MA", county="Suffolk", tract=070200, check=T)
Testing geography item 1: Tract 70200, Suffolk County, Massachusetts .... OK.
> tract.2=geo.make(state="MA", county="Suffolk", tract=070402, check=T)
Testing geography item 1: Tract 70402, Suffolk County, Massachusetts .... OK.
> chinatown=geo.make(state=25, county=25, tract=c(70200,70402))

Thanks for your help and many many thanks for developing this package which has made my life so much easier.

Comment on Major improvements to acs.R: sneak peak at version 1.0 by John

John — Sun, 24 Feb 2013 22:05:44 +0000

Not to be that guy, but this is an amazing prospect for those of us who work with Census and ACS data, so can we get a more concrete timeline on this? When it’s going to be done? And what “done” means – how many ACS years, data sets, etc its going to cover? Those of us who work with Census data are immensely grateful for all your hard work!

Comment on acs Package Updated: version 0.8 now on CRAN by Ezra Glenn

Ezra Glenn — Wed, 13 Feb 2013 01:10:07 +0000

Update: see the added download functionality in the new version of the package: http://eglenn.scripts.mit.edu/citystate/2013/02/major-improvements-to-acs-r-sneak-peak-at-version-1-0/

Comment on Major improvements to acs.R: sneak peak at version 1.0 by Ezra Glenn

Ezra Glenn — Tue, 12 Feb 2013 16:43:39 +0000

Mara:

Thanks for the comment/question. As for which version, you should be
fine with almost anything recent (and certainly R version 2.15.2,
which it seems you are running). My guess is that you tried to
install the package from the CRAN repository, rather than downloading
the “acs_1.0.tar.gz” file from the CityState website
(http://eglenn.scripts.mit.edu/citystate/2013/02/major-improvements-to-acs-r-sneak-peak-at-version-1-0/).
Since this is a pre-release version, installation is a bit different.

After you download the file and save it on your computer somewhere,
you start R and type:

> install.packages(“path/to/acs_1.0.tar.gz”)

If that doesn’t work, try these additional options:

> install.packages(“/path/to/acs_1.0.tar.gz”, repos = NULL, type = “source”)

One of those should work, and then you’re good to go — but see the
draft user guide, too:
< http://eglenn.scripts.mit.edu/citystate/wp-content/uploads/2013/02/wpid-working_with_acs_R2.pdf>.

Once we catch all the bugs, I’ll add this to the repository, and you
will be able to install with just:

> install.packages(“acs”)

but for now, you need these extra steps.

Let me know how it goes!

Comment on Major improvements to acs.R: sneak peak at version 1.0 by Mara Averick

Mara Averick — Tue, 12 Feb 2013 16:31:00 +0000

First off, thank you for this amazing tool- hopefully this will save me from having that extra step between data retrieval from the ACS API and importing it into R.

Seemingly stupid question- what version of R should I be using?

Warning message: package ‘acs_1.0.tar.gz’ is not available (for R version 2.15.2)

Comment on acs Package Updated: version 0.8 now on CRAN by Peter

Peter — Thu, 24 May 2012 15:54:30 +0000

Great package, and one that I intend to make use of! Have you considered instrumenting this package to automate downloads of the appropriate ACS data programmatically from R? I looked into how to fetch the Census data in an automated fashion, but was dismayed to find that the American Factfinder2's Deep Linking API only gets you as far as a webpage, rife with javascript controls, from which one *could* download the appropriate tables. I know that the USCensus2000 package provides some facilities for grabbing a whole lot of Census data from R, however my guess is that the decennial Census data is an easier automation problem.

Comment on Hurdy Gurdy (Daniel Seideneder and Daniel Pfeiffer, 2011) by Daniel Pfeiffer

Daniel Pfeiffer — Tue, 08 May 2012 14:26:22 +0000

Dear Mr. Glenn,
thank you very much for review, critic and compliments for our film. We are very happy to get such a good feedback about Hurdy Gurdy.
Carlos asked about the Musik. It is a new composition, especially made for our film. The title is Hurdy Gurdy. The composer is Ben Hansen.
We shoot the Film in Estonia with Tilt/shift lenses (real world) and in Hamburg at Wunderwelt (Model World). So we combined the real world with a model world. We try to connect both worlds, with editing and color correction, so that the difference is hardly recognizable.
If you have any questions or what ever, it would please me, when you contact me.

Comment on org2blog by Jason John Wells

Jason John Wells — Sun, 06 May 2012 17:31:16 +0000

Thanks for the tip. I think I overlooked a static HTML front end. I should probably keep it simple as tinkering and comparing a “better solution” can in the way of actually publishing something.

Have you come across Kurt Schwehr’s Research Tools course material? He has published some great .org class lecture note files and YouTube videos.

He covers Emacs, Org, Python and GIS.

http://schwehr.blogspot.jp/2011/10/research-tools-lecture-1-introduction.html