rs_z: Shiller's Z matrix
In marberts/ppd: Prices Production and Development

Description Usage Arguments Details References See Also Examples

Compute the 'Z' matrix in Shiller (1991, section I) from sales-pair data to calculate a repeat-sales price index.

1	rs_z(x, y, f, sparse = FALSE)

`x, y`	Period of the second and first sale, respectively. Usually a vector of dates.
`f`	An optional factor the same length as `x` and `y`, or an atomic vector to be turned into a factor, that is used to group `x` and `y`.
`sparse`	Should sparse matrices from Matrix be used in the calculation (faster for large datasets), or regular dense matrices?

The Z matrix is not well defined if either x or y have missing values, and an error is thrown if any inputs have missing values.

Shiller, R. J. (1991). Arithmetic repeat sales price estimators. Journal of Housing Economics, 1(1):110-126.

See rs_x for making Shiller's X matrix.

# Make some data
x <- data.frame(date = c(3, 2, 3), date_prev = c(1, 1, 2), price = 3:1, price_prev = 1)

# Calculate the GRS index
Z <- rs_z(x$date, x$date_prev)
y <- log(x$price) - log(x$price_prev)
b <- solve(crossprod(Z[, -1]), crossprod(Z[, -1], y))[, 1]
exp(b) * 100