SetDynamicCovariates: Add Dynamic Covariates to a CLV data object

In bachmannpatrick/CLVTools: Tools for Customer Lifetime Value Estimation

Add Dynamic Covariates to a CLV data object

Description

Add dynamic covariate data to an existing data object of class clv.data. The returned object can be used to fit models with dynamic covariates.

No covariate data can be added to a clv data object which already has any covariate set.

At least 1 covariate is needed for both processes and no categorical covariate may be of only a single category.

Usage

SetDynamicCovariates(
  clv.data,
  data.cov.life,
  data.cov.trans,
  names.cov.life,
  names.cov.trans,
  name.id = "Id",
  name.date = "Date"
)

Arguments

clv.data	CLV data object to add the covariates data to.
data.cov.life	Dynamic covariate data as data.frame or data.table for the lifetime process.
data.cov.trans	Dynamic covariate data as data.frame or data.table for the transaction process.
names.cov.life	Vector with names of the columns in data.cov.life that contain the covariates.
names.cov.trans	Vector with names of the columns in data.cov.trans that contain the covariates.
name.id	Name of the column to find the Id data for both, data.cov.life and data.cov.trans.
name.date	Name of the column to find the Date data for both, data.cov.life and data.cov.trans.

Details

data.cov.life and data.cov.trans are data.frames or data.tables that each contain exactly 1 row for every combination of timepoint and customer. For each customer appearing in the transaction data there needs to be covariate data at every timepoint that marks the start of a period as defined by time.unit. It has to range from the start of the estimation sample (timepoint.estimation.start) until the end of the period in which the end of the holdout sample (timepoint.holdout.end) falls. See the the provided data apparelDynCov for illustration. Covariates of class character or factor are converted to k-1 numeric dummies.

Date as character If the Date column in the covariate data is of type character, the date.format given when creating the the clv.data object is used for parsing.

Value

An object of class clv.data.dynamic.covariates. See the class definition clv.data.dynamic.covariates for more details about the returned object.

Examples

## Not run: 
data("apparelTrans")
data("apparelDynCov")

# Create a clv data object without covariates
clv.data.apparel <- clvdata(apparelTrans, time.unit="w",
                            date.format="ymd")

# Add dynamic covariate data
clv.data.dyn.cov  <-
   SetDynamicCovariates(clv.data.apparel,
                       data.cov.life  = apparelDynCov,
                       names.cov.life = c("Marketing", "Gender", "Channel"),
                       data.cov.trans = apparelDynCov,
                       names.cov.trans = c("Marketing", "Gender", "Channel"),
                       name.id = "Id",
                       name.date = "Cov.Date")

# summary output about covariates data
summary(clv.data.dyn.cov)

# fit pnbd model with dynamic covariates
pnbd(clv.data.dyn.cov)

## End(Not run)

bachmannpatrick/CLVTools documentation built on Oct. 29, 2023, 2:16 p.m.