explore_corr: Explore correlations

In realgabon/insuRglm: Tools for GLM Modeling in Insurance Context

Explore correlations

Description

Helps to explore associations (correlations between categorical variables) in the data or in the current (last) insuRglm model. It can be either through visual or tabular form. The metric used for strength of association is Cramer's V.

Usage

explore_corr(setup, type = c("visual", "tabular"))

Arguments

setup	Setup object. Created at the start of the workflow. Usually piped in from previous step.
type	Character scalar. Either visual or tabular.

Value

Either a dataframe or a ggplot2 chart.

See Also

explore_data, explore_target

Examples

require(dplyr) # for the pipe operator
data('sev_train')

setup <- setup(
  data_train = sev_train,
  target = 'sev',
  weight = 'numclaims',
  family = 'gamma',
  keep_cols = c('pol_nbr', 'exposure', 'premium')
)

explore_corr(setup, type = 'visual')
explore_corr(setup, type = 'tabular')

modeling <- setup %>%
  factor_add(pol_yr) %>%
  factor_add(agecat) %>%
  factor_add(veh_age) %>%
  factor_add(veh_value) %>%
  model_fit()

explore_corr(modeling, type = 'visual')
explore_corr(modeling, type = 'tabular')

realgabon/insuRglm documentation built on Jan. 2, 2023, 2:51 a.m.