sample_transactions: Sample Transactions

In ben519/trinalysis: Transaction Data Analysis

Description

Generate a sample transactions dataset such that the simulated data has a (somewhat) realistic pattern of customer churn

Usage

sample_transactions(nCusts, sdTransactions = 10, sdAmount = 10,
  minDate = as.Date("2014-1-1"), maxDate = as.Date("2015-12-31"),
  products = LETTERS)

Arguments

nCusts	How many customers?
sdTransactions	The number of transactions per customer = abs(ceiling(rnorm(n=1, sd=sdTransactions)))
sdAmount	The amount of each transaction = abs(round(rnorm(n=1, sd=sdAmount),2))
minDate	The minimum possible date a transaction could occur
maxDate	The maximum possible date a transaction could occur
products	A character vector of potential product names

Details

Returns a data.table object with columns TransactionID, TransactionDate, CustomerID, Amount, Product

Examples

library(data.table)

set.seed(2016)
sample_transactions(nCusts=10)
sample_transactions(nCusts=10, sdTransactions=5, sdAmount=1000)
sample_transactions(nCusts=10, minDate=as.Date("2000-1-1"), maxDate=as.Date("2016-12-31"))
sample_transactions(nCusts=10, products=c("baseball", "football", "soccerball"))

ben519/trinalysis documentation built on May 12, 2019, 10:56 a.m.