18 February 2012

How-to Extract Text From Multiple Websites with R

I have been meaning to post this slideshow for awhile now. It gives a brief introduction to using R for scraping text from multiple websites. It includes some basic debugging, because R sometimes misses a website.

Just click the arrows to change the slides. Enjoy!

8 February 2012

OpenCPU, R in the Cloud

I ran across OpenCPU today. If you have any intest in R and reproducible research this is definitely worth checking out. Also, it looks like I might want to explore the potential of embedding functions in websites. Hm . . . .

7 February 2012

Update to Partisan Bias in Fed Inflation Forecasts

Since I'm in the depths of PhD thesis revisions I haven't had much time to do much other than update previous posts (see my Stata Country Standardizer Update).

Here is an update of an earlier post about possible partisan biases in US Federal Reserve staff inflation forecasts (these influence Federal Open Market Committee meetings where US monetary policy is largely made). The new graph below allows us to see even more of what has been going on over time.

The partisan effect is less obvious than in the earlier graph, but is is clear that during this time period the big over estimations are during Democratic presidencies and the big (actually almost all) underestimations are during Republican ones. The effect would be even stronger if we took out the end of Reagan's first term and his second one, where Fed staff may not have fully adjusted their forecasting to reflect the Volker-Greenspan era of moderate inflation.

For more details about the graph (sources, how 'error' is defined, etc.) see the earlier post.

Greenbook Inflation Forecast Errors, by Presidential Party
Note, the shaded area indicates minimal error.

The R Code: