Skip to contents

Predict injuries

Source: R/injuries_function_2.R
injuries_function_2.Rd

Predict injuries for baseline and scenarios based on Poisson regression model fitted on baseline fatality counts and distances

Usage

injuries_function_2(
  true_distances,
  injuries_list,
  reg_model,
  constant_mode = F
)

Arguments

true_distances

data frame containing population distances for each scenario

injuries_list

list of dataframes set up with scenario specific information to supply to regression model for prediction

reg_model

Poisson injury regression model

constant_mode

whether or not we are in constant (vs sampling) mode

Value

injuries2 - dataframe containing predicted fatality counts for each casualty mode by age and sex and for each scenario, plus confidence interval limits for constant mode

whw_temp - list containing the fatality predictions for each casualty and strike mode pair split into whw and nov matrices for each scenario. Upper and lower confidence interval predictions are also included for the constant mode

Details

This function uses the Poisson regression model built in the distances_for_injury_function() to predict fatality counts for the Baseline and all the scenarios. It performs the following steps:

  • create an injuries data frame containing all the distances travelled by mode, age, sex and scenario

  • predict the fatalities for each strike and casualty mode combination, age and sex category and each scenario (incl Baseline). If the sample mode is set to 'constant' (and not 'sample'), we also predict upper and lower confidence interval boundaries

  • if running in constant mode, create a whw_temp list containing the total predicted fatality counts for each casualty and strike mode pair for each scenario split into whw and nov matrices and also give the upper and lower confidence interval limit predictions. Also create a combined outcome table where NOV fatalities are added as casualty mode equals strike mode fatalities.

  • create an injuries2 data frame containing the total predicted fatality counts for each casualty mode by age and sex for each scenario. This dataframe also contains total death per age and sex category and, for the constant mode the upper and lower total death predictions of the confidence interval.