cusum(x, ...)
ccusum(x,mu=NULL, k=0.5,h=4, ...)
bcusum(x,prob=NULL, R0=1.0,Ra=2, ...)
R0=1
, Ra=2
To detect increases we set Ra > R0
, otherwise Ra < R0
.
In bcusum
, the probability c0 under H0 is computed based on
current data, which can also be given as a parameter prob
.
Choose lower and upper limits h0 and h1: h1=1.5, 2.0, 2.5, 3, 10; h0=-2.5
Wald: h0 = -ln(1-alpha)/beta, h1 = ln(1-beta)/alpha. Alert lines: alpha=beta=0.1. Alarm/action lines: alpha=beta=0.01.
Rogers et al defined a scale adjustment ln(OR), say log(1.5)
de Leval et al: alert line with alpha=0.05, beta=0.2; alarm line with alpha=0.01, beta=0.1
Page, E. S., 1954. Continuous inspection schemes. Biometrika 41 (1/2), pp. 100–115. URL http://www.jstor.org/stable/2333009
Parsonnet, V., Dean, D., Bernstein, A. D., Jun 1989. A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 79, 3–12.
Steiner, S. H., Cook, R. J., Farewell, V. T., Jan 1999. Monitoring paired binary surgical outcomes using cumulative sum charts. Stat Med 18, 69–86.
Steiner, S. H., Cook, R. J., Farewell, V. T., Treasure, T., Dec 2000. Monitoring surgical performance using risk-adjusted cumulative sum charts. Biostatistics 1, 441–452.
robot
data(robot)
hst = robot$TMHSTMIN
out = cusum(hst,mu=45)
plot(out)
Run the code above in your browser using DataLab