pricelevels (version 1.4.0)

ratios: Calculation of price ratios

Description

Calculation of regional price ratios per product with flexible setting of base prices.

Usage

ratios(p, r, n, q=NULL, w=NULL, base=NULL, settings=list())

Value

A numeric vector of the same length as p. If base is not NULL and static=FALSE, the attribute attr(x, "base") is added to the output, providing the respective base region for each product.

Arguments

p

A numeric vector of positive prices.

r, n

A character vector or factor of regional entities r and products n, respectively.

q, w

A numeric vector of non-negative quantities q or expenditure share weights w. If both q and w are provided, q will be used. This is only relevant for the averaging of duplicated prices (see Section 'Details').

base

A character specifying the base region to be used for the calculation of price ratios. If NULL, price ratios are calculated with reference to the regional average price of a product (see Section 'Details')

settings

A list of control settings to be used. The following settings are supported:

• chatty : A logical specifying if warnings and info messages should be printed or not. The default is getOption("pricelevels.chatty").

• static : A logical indicating whether the base region is static (TRUE), that is, always the same, or if another region than base is allowed to be used when prices for base are not available or NA for specific products. Only relevant if base is not NULL. The default is TRUE.

Author

Sebastian Weinand

Details

If there are \(k=1,\ldots,K^{r}_{n}\) duplicated prices for product \(n\) in region \(r\), these are averaged using the quantities q (or similarly as a weighted arithmetic mean using the weights w) if provided:

$$\bar{p}^{r}_n = \sum_{k=1}^{K^r_n} p^r_{nk} q^r_{nk} \Big/ \sum_{k=1}^{K^r_n} q^r_{nk}$$

Price ratios are then derived for each product \(n\) by \(p_n^r \Big/ \frac{1}{R} \sum_{s=1}^{R} \bar{p}_{n}^s\) if base=NULL and by \(p_n^r \big/ \bar{p}^{base}_n\) otherwise.

Examples

### (1) unique price observations

set.seed(123)
dt1 <- rdata(R=3, B=1, N=4)
levels(dt1$region) <- c("a","b","c")

# calculate price ratios by product:
dt1[, ratios(p=price, r=region, n=product, base="b")]


### (2) unique price observations and missing base region


# drop two observations:
dt2 <- dt1[-c(5,10), ]

# now, region 'a' is base for product 2:
dt2[, "pr" := ratios(p=price, r=region, n=product, base="b",
                     settings=list(static=FALSE))]

# base regions are stored in attributes:
attr(dt2$pr, "base")

# with static base, NAs are produced:
dt2[, "pr_static" := ratios(p=price, r=region, n=product, base="b")]


### (3) duplicated prices

# insert duplicates and missings:
dt3 <- rbind(dt1[c(2,3),], dt1[-c(11),])
dt3[1:2, c("price","quantity") := list(price*1.1, quantity*0.95)]
anyDuplicated(dt3, by=c("region","product"))

# duplicated prices are divided by the weighted average base prices:
dt3[, ratios(p=price, r=region, n=product, q=quantity, base="b",
             settings=list(static=FALSE))]