Closed pastephens closed 12 years ago
darribas
commented
12 years ago Related to this, in case we need it, the R library to access and visualize the changing boundaries of the world countries is cshapes (CRAN link), and is described in this website and this paper in the R Journal.
Maybe we can use these as the spatial data to link it to Johnny's database. Other than that, it's really fun to play with :-)
geoquant
commented
12 years ago Came across this website with a substantial amount of historical data sets:
pastephens
commented
12 years ago Great source of interesting data!
"Rice prices in 14 regions, 1620-1867" has 14 spatial units and 247 temporal observations, with the bonus complication of staggered series start dates, but full arrays once they start. http://gpih.ucdavis.edu/files/Japan_rice_P_1620-1867.xls
"Prices of various goods in Belgian towns during the 1870s" has just 4 time periods, about 50 cities in 9 countries, for about 20 different commodity prices. http://gpih.ucdavis.edu/files/Ward-Devereux_P_1872-78.xls
"California wages, 1870-1928" includes the table above it on SF wholesale prices, making it a dataset with which one could attempt to gauge the standard of living available to different types of workers. The wage data includes the average work week and the average dollars per day for a variety of occupations. http://gpih.ucdavis.edu/files/San_Francisco_wholesale_prices_1847-1900_and_California_wages_1870-1928.xls
The idea expressed at our last meeting (1/24/2012) was to have 3 data sets with different space-time properties: one set with large spatial cross-section (n) and 3 time series (k), one set with a small spatial cross-section and many time series, and a third set with large n and large k.