orangeJuice: Store-level Panel Data on Orange Juice Sales

In bayesm: Bayesian Inference for Marketing/Micro-Econometrics

Store-level Panel Data on Orange Juice Sales

Description

Weekly sales of refrigerated orange juice at 83 stores. Contains demographic information on those stores.

Usage

data(orangeJuice)

Format

The orangeJuice object is a list containing two data frames, yx and storedemo.

Details

In the yx data frame:

...$store	store number
...$brand	brand indicator
...$week	week number
...$logmove	log of the number of units sold
...$constant	a vector of 1s
...$price#	price of brand #
...$deal	in-store coupon activity
...$feature	feature advertisement
...$profit	profit

The price variables correspond to the following brands:

1	Tropicana Premium 64 oz
2	Tropicana Premium 96 oz
3	Florida's Natural 64 oz
4	Tropicana 64 oz
5	Minute Maid 64 oz
6	Minute Maid 96 oz
7	Citrus Hill 64 oz
8	Tree Fresh 64 oz
9	Florida Gold 64 oz
10	Dominicks 64 oz
11	Dominicks 128 oz

In the storedemo data frame:

...$STORE	store number
...$AGE60	percentage of the population that is aged 60 or older
...$EDUC	percentage of the population that has a college degree
...$ETHNIC	percent of the population that is black or Hispanic
...$INCOME	median income
...$HHLARGE	percentage of households with 5 or more persons
...$WORKWOM	percentage of women with full-time jobs
...$HVAL150	percentage of households worth more than $150,000
...$SSTRDIST	distance to the nearest warehouse store
...$SSTRVOL	ratio of sales of this store to the nearest warehouse store
...$CPDIST5	average distance in miles to the nearest 5 supermarkets
...$CPWVOL5	ratio of sales of this store to the average of the nearest five stores

Source

Alan L. Montgomery (1997), "Creating Micro-Marketing Pricing Strategies Using Supermarket Scanner Data," Marketing Science 16(4) 315–337.

References

Chapter 5, Bayesian Statistics and Marketing by Rossi, Allenby, and McCulloch.

Examples

## load data
data(orangeJuice)

## print some quantiles of yx data  
cat("Quantiles of the Variables in yx data",fill=TRUE)
mat = apply(as.matrix(orangeJuice$yx), 2, quantile)
print(mat)

## print some quantiles of storedemo data
cat("Quantiles of the Variables in storedemo data",fill=TRUE)
mat = apply(as.matrix(orangeJuice$storedemo), 2, quantile)
print(mat)


## processing for use with rhierLinearModel
if(0) {
  
  ## select brand 1 for analysis
  brand1 = orangeJuice$yx[(orangeJuice$yx$brand==1),]
  
  store = sort(unique(brand1$store))
  nreg = length(store)
  nvar = 14
  
  regdata=NULL
  for (reg in 1:nreg) {
    y = brand1$logmove[brand1$store==store[reg]]
    iota = c(rep(1,length(y)))
    X = cbind(iota,log(brand1$price1[brand1$store==store[reg]]),
                   log(brand1$price2[brand1$store==store[reg]]),
                   log(brand1$price3[brand1$store==store[reg]]),
                   log(brand1$price4[brand1$store==store[reg]]),
                   log(brand1$price5[brand1$store==store[reg]]),
                   log(brand1$price6[brand1$store==store[reg]]),
                   log(brand1$price7[brand1$store==store[reg]]),
                   log(brand1$price8[brand1$store==store[reg]]),
                   log(brand1$price9[brand1$store==store[reg]]),
                   log(brand1$price10[brand1$store==store[reg]]),
                   log(brand1$price11[brand1$store==store[reg]]),
                   brand1$deal[brand1$store==store[reg]],
                   brand1$feat[brand1$store==store[reg]] )
    regdata[[reg]] = list(y=y, X=X)
    }
  
  ## storedemo is standardized to zero mean.
  Z = as.matrix(orangeJuice$storedemo[,2:12]) 
  dmean = apply(Z, 2, mean)
  for (s in 1:nreg) {Z[s,] = Z[s,] - dmean}
  iotaz = c(rep(1,nrow(Z)))
  Z = cbind(iotaz, Z)
  nz = ncol(Z)
  
  Data = list(regdata=regdata, Z=Z)
  Mcmc = list(R=R, keep=1)
  
  out = rhierLinearModel(Data=Data, Mcmc=Mcmc)
  
  summary(out$Deltadraw)
  summary(out$Vbetadraw)
  
  ## plotting examples
  if(0){ plot(out$betadraw) }
}

bayesm documentation built on Sept. 24, 2023, 1:07 a.m.