TPD_decomp: Decomposition of Fixed-Effects index
In MjStansfi/TPDdecomp:

Description Usage Arguments Value Examples

View source: R/TPD_decomp.R

Take (mostly) the same inputs as the FEWS function and return the decomposition on the multilateral scale.

1 2	TPD_decomp(times, logprice, id, weight, custom_time = c(), window_length = NULL, verbose = FALSE)

`times`	vector of the times at which price observations were made. NOTE: `times` must be of class Date or numeric.
`logprice`	vector of log of prices at the given time
`id`	vector of distinct identification number of consumer goods
`weight`	vector of expenditure weights used in the regressions
`custom_time`	either empty (will assume latest periods) or a vector of length two relating to times compared where [1] is the 'from' time and [2] is the 'to' time e.g Comparing contribution from "1973-04-01" to "1973-05-01" would be c("1973-04-01", "1973-05-01")
`window_length`	optional. Single number for length of window for the data that regressions are fit on. Note if window_length is present it assumes custom time is latest two periods in window.

A dataframe containing the numeric contribution (contrib) of every price observation in the window, the product of which will equal the index movement between the two time periods.

p_contrib is the standardised percentage contribution of the observation, the sum of which will equal 1. id_p_contrib is the standardised percentage contribution of the total id, the sum of which will equal 1.

library(TPDdecomp)

load("Turvey.RData")

contributions <- with(
    turvey,
    TPD_decomp(times = month,
               logprice = log(price),
               id = commodity,
               weight = price*quantity,
               custom_time = c("1973-04-30","1973-05-31"),
               window_length = NULL)
 )
#Comparing 1973-05-31 to 1973-04-30

str(contributions)
Classes ‘data.table’ and 'data.frame':	176 obs. of  8 variables:
$ times       : Date, format: "1973-07-31" "1973-06-30" "1973-05-31" ...
$ price       : num  5.19 4.53 7.17 4.3 3.98 6.21 3.78 3.72 5.68 3.27 ...
$ weight      : num  14932 18505 7407 10892 14495 ...
$ id          : chr  "Strawberries" "Strawberries" "Strawberries" "Strawberries" ...
$ exp_share   : num  0.36 0.413 0.238 0.342 0.41 ...
$ contrib     : num  0.965 0.967 1.545 0.97 0.97 ...
$ p_contrib   : num  -0.208 -0.201 2.58 -0.181 -0.184 ...
$ id_p_contrib: num  0.663 0.663 0.663 0.663 0.663 ...