treatments_by_policy: Use sim_treatment_by_subgroup to simulate treatments

Description Usage Arguments Value Examples

View source: R/treatment.R

Description

Simulate treatments according to specified policy rules

Usage

1
treatments_by_policy(policies, treat_chars, stagegroups, map, pop_size, nsim)

Arguments

policies

A "scenarios" data frame containing an 'id' for the policies and a 'pairnum' column indicating either NA or the paired policy ID, for strategies with early detection. See ex1$pol

treat_chars

Data frame with "txSSno" column indicating treatment numbers and subsequent columns with treatment proportions WITHIN stage-subgroups. Each of these columns should correspond to a row in the "policies" data frame, with their names taken fro policies$id. See ex1$tx

stagegroups

List of stage-subgroup matrices, one for each policy/row in the "scenarios" data frame

map

Stage-subgroup map indicating allowed stage-shifts. See ex1$map.

pop_size

Population size (number of rows)

nsim

Number of sims (number of columns)

Value

List of treatment matrices, one for each policy in the "scenarios" data frame. Each matrix contains treatment IDs corresponding to treat_chars$txSSno. Early detection scenarios will have NAs for advanced-stage cases who aren't stage-shifted.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
library(bcimodel)
data(ex1) 
# ex1$nh shows that there are 4 stage-subgroups. Use a fake random distribution of groups 1:4 for the population before stage-shifting.
popdistr <- matrix(sample.int(4, size=40, replace=TRUE), nrow=20, ncol=2)
# Create stageshift indicator matrices for all 3 scenarios: no stage shifts for #1 and #2, but 30% stageshift for #3. Use a small population of size 20, and 2 sims
stageshifts <- list(base=matrix(0, nrow=20, ncol=2),
                    tam=matrix(0, nrow=20, ncol=2), 
                    tamshift=stageshift_indicator(0.85, 20, 2))
# Get the actual stages - only policy #3 has stage-shifting
stages <- lapply(stageshifts, shift_stages, original=popdistr, map=ex1$map)
lapply(stages, table)
t <- treatments_by_policy(policies=ex1$pol, 
                          treat_chars=ex1$tx, 
                          stagegroups=stages, 
                          map=ex1$map, 
                          pop_size=20, nsim=2)

cancerpolicy/bcimodel documentation built on Feb. 13, 2018, 2:06 p.m.