CLVTools (version 0.11.2)

latentAttrition: Formula Interface for Latent Attrition Models

Description

Fit latent attrition models for transaction behavior, using a formula to specify the covariates.

Usage

latentAttrition(
  formula,
  family,
  data,
  optimx.args = list(),
  verbose = TRUE,
  ...
)

Arguments

formula

Formula to select and transform covariates in data. Has to be left empty if data contains no covariates. See Details.

family

A latentAttrition model. One of pnbd, bgnbd, or ggomnbd.

data

A clv.data object.

optimx.args

Additional arguments to control the optimization which are forwarded to optimx::optimx. If multiple optimization methods are specified, only the result of the last method is further processed.

verbose

Show details about the running of the function.

...

Forwarded to model specified in family.

Details

A two-part formula is used to select and transform the covariates stored in data before the model is estimated on it. May not be given if data contains no covariates.

The formula left hand side (LHS) has to remain empty and may never be specified.

The formula right hand side (RHS) follows a two-part notation using | as separator.

• 1st part: Which covariates to include for the lifetime process, potentially transforming them and adding interactions. The dot ('.') refers to all lifetime covariates.

• 2nd part: Which covariates to include for the transaction process, potentially transforming them and adding interactions. The dot ('.') refers to all transaction covariates

e.g: ~ covlife | covtrans

See the example section for illustrations on how to specify the formula parameter.

See Also

Models for inputs to family: pnbd, ggomnbd, bgnbd.

spending to fit spending models with a formula interface

Examples

# \donttest{

data("apparelTrans")
data("apparelStaticCov")

clv.nocov <-
    clvdata(apparelTrans, time.unit="w", date.format="ymd")

# Create static covariate data with 2 covariates
clv.staticcov  <-
  SetStaticCovariates(clv.nocov,
                      data.cov.life  = apparelStaticCov,
                      names.cov.life = c("Gender", "Channel"),
                      data.cov.trans = apparelStaticCov,
                      names.cov.trans = c("Gender", "Channel"))

# Fit models without covariates.
# Note that NO formula may be specified in this case
latentAttrition(formula =, family=pnbd, data=clv.nocov)
latentAttrition(formula =, family=bgnbd, data=clv.nocov)
latentAttrition(formula =, family=ggomnbd, data=clv.nocov)

# Fit pnbd with start parameters and correlation
# required args are passed as part of '...'
latentAttrition(formula =, family=pnbd, data=clv.nocov,
                use.cor=TRUE,
                start.params.model=c(r=1, alpha=10, s=2, beta=8))


# Fit pnbd with all present covariates
latentAttrition(formula=~.|., family=pnbd, data=clv.staticcov)

# Fit pnbd with selected covariates
latentAttrition(formula=~Gender|Channel+Gender, family=pnbd,
                data=clv.staticcov)

# Fit pnbd with start parameters for covariates
latentAttrition(formula=~Gender|., family=pnbd,
                data=clv.staticcov,
                start.params.life = c(Gender = 0.6),
                start.params.trans = c(Gender = 0.6, Channel = 0.4))

# Fit pnbd with transformed covariate data
latentAttrition(formula=~Gender|I(log(Channel+2)), family=pnbd,
                data=clv.staticcov)

# Fit pnbd with all covs and regularization
latentAttrition(formula=~.|., family=pnbd, data=clv.staticcov,
                reg.lambdas = c(life=3, trans=8))

# Fit pnbd with all covs and constraint parameters for Channel
latentAttrition(formula=~.|., family=pnbd, data=clv.staticcov,
                names.cov.constr='Channel')

# }