Corrected Boston Housing Data

Description

The boston.c data frame has 506 rows and 20 columns. It contains the Harrison and Rubinfeld (1978) data corrected for a few minor errors and augmented with the latitude and longitude of the observations. Gilley and Pace also point out that MEDV is censored, in that median values at or over USD 50,000 are set to USD 50,000. The original data set without the corrections is also included in package mlbench as BostonHousing. In addition the following objects are included: a matrix of tract point coordinates projected to UTM zone 19 (included as boston.utm), and a sphere of influence neighbours list (boston.soi) and the one using corrected values in the package spdep.

The data frame dataBoston contains the variables CMEDV, DIS, RAD and LSTAT already in logarithms. On top of that squares of the regressors RM (RM#RM) and NOX (NOX#NOX) are provided to model potential non-linearities in the data. Furthermore, WL.boston is a list object containing extracted eigenvectors corresponding to five different spatial weight matrices (WL.boston: the boston.soi matrix, two random perturbations of the boston.soi matrix and a knn-4 and knn-6 matrix.). The original matrices are contained in the list object weightM.boston

Usage

data(dataBoston)

Format

A data frame with 506 observations on the following variables.

TOWN

a factor with levels given by town names

TOWNNO

a numeric vector corresponding to TOWN

TRACT

a numeric vector of tract ID numbers

LON

a numeric vector of tract point longitudes in decimal degrees

LAT

a numeric vector of tract point latitudes in decimal degrees

MEDV

a numeric vector of median values of owner-occupied housing in USD 1000

CMEDV

a numeric vector of corrected median values of owner-occupied housing in USD 1000

CRIM

a numeric vector of per capita crime

ZN

a numeric vector of proportions of residential land zoned for lots over 25000 sq. ft per town (constant for all Boston tracts)

INDUS

a numeric vector of proportions of non-retail business acres per town (constant for all Boston tracts)

CHAS

a factor with levels 1 if tract borders Charles River; 0 otherwise

NOX

a numeric vector of nitric oxides concentration (parts per 10 million) per town

RM

a numeric vector of average numbers of rooms per dwelling

AGE

a numeric vector of proportions of owner-occupied units built prior to 1940

DIS

a numeric vector of weighted distances to five Boston employment centres

RAD

a numeric vector of an index of accessibility to radial highways per town (constant for all Boston tracts)

TAX

a numeric vector full-value property-tax rate per USD 10,000 per town (constant for all Boston tracts)

PTRATIO

a numeric vector of pupil-teacher ratios per town (constant for all Boston tracts)

B

a numeric vector of 1000*(Bk - 0.63)^2 where Bk is the proportion of Afro-Americans

LSTAT

a numeric vector of percentage values of lower status population

Source

http://lib.stat.cmu.edu/datasets/boston_corrected.txt

References

Harrison, David, and Daniel L. Rubinfeld, Hedonic Housing Prices and the Demand for Clean Air, Journal of Environmental Economics and Management, Volume 5, (1978), 81-102. Original data.

Gilley, O.W., and R. Kelley Pace, On the Harrison and Rubinfeld Data, Journal of Environmental Economics and Management, 31 (1996), 403-405. Provided corrections and examined censoring.

Pace, R. Kelley, and O.W. Gilley, Using the Spatial Configuration of the Data to Improve Estimation, Journal of the Real Estate Finance and Economics, 14 (1997), 333-340.

Examples

data(dataBoston)