nc_retain: Predict retention across a set of ages for NC methodology
In southwick-associates/lifetime: Provide functions for lifetime pricing

Description Usage Arguments Functions See Also Examples

Convenience functions for the NC analysis, mainly wrappers for yrs_result_retain

1
2
3

nc_retain(retain_all)

nc_retain_all(history_split, ages, use_observed = FALSE)

`retain_all`	full set of retention curve results produced by nc_retain_all()
`history_split`	license history list produced by yrs_zero_split()
`ages`	set of ages for which retention curves will be calculated separately
`use_observed`	if TRUE, return observed (instead of predicted) retention rates

nc_retain_all: Predict full retention curves

Other wrapper functions for NC results: nc_break_even_yrs, nc_break_even, nc_price_lifetime_youth, nc_retain_youth, nc_revenue

library(dplyr)
library(ggplot2)
data(all_sports, lic, sale)

df_split <- all_sports %>%
    yrs_lifetime_join(sale, lic, "sportsman") %>%
    yrs_zero_split() %>%
    yrs_zero_filter(function(x) filter(x, life_group == "sportsman"))

observe_all <- nc_retain_all(df_split, 52:58, use_observed = TRUE)
retain_all <- nc_retain_all(df_split, 52:58)
retain <- nc_retain(retain_all)
retain %>%
    ggplot(aes(current_age, yrs)) +
    geom_point() +
    ggtitle("Predicted years of purchases by current age")

# observed vs predicted for specified ages
observe <- filter(observe_all, current_age == 52)
retain <- filter(retain_all, current_age == 52)
yrs_plot(retain) + geom_point(data = observe) +
    ggtitle("Some noisy data points due to sample size")

observe <- filter(observe_all, current_age == 58)
retain <- filter(retain_all, current_age == 58)
yrs_plot(retain) + geom_point(data = observe) +
    ggtitle("Observed rates are used for old enough buyers",
            "Senior lifetime artifacts emerge at age 65")