Exercise

# Logistic regression with 2 explanatory variables

To include multiple explanatory variables in logistic regression models, the syntax is the same as for linear regressions. The only change is the same as in the simple case: you run a *generalized* linear model with a binomial error family.

Here you'll fit a model of churn status with both of the explanatory variables from the dataset: the length of customer relationship and the recency of purchase.

`churn`

is available.

Instructions

**100 XP**

- Fit a logistic regression of churn status,
`has_churned`

versus length of customer relationship,`time_since_first_purchase`

and recency of purchase,`time_since_last_purchase`

, and an interaction between the explanatory variables.