nordpred: Fit power5 and poisson Age-Period-Cohort models for...

Description Usage Arguments Details Value Note for S-plus Author(s) References See Also Examples

Description

nordpred uses the power5 and poisson Age-Period-Cohort (APC) models to calculate prediction of cancer incidence and mortality.

Usage

1
2
3
4
5
nordpred(cases, pyr, startestage, startuseage, noperiods = NULL,
  recent = NULL, cuttrend = c(0, 0.25, 0.5, 0.75, 0.75),
  linkfunc = "power5")

is.nordpred(x)

Arguments

cases

A data.frame with number of cases

pyr

A data.frame with observed and forecasted person years.

startestage

Youngest age group to be included in the regression model. Predictions for age groups below this limit it based on average rates from the last 10 years.

startuseage

Youngest age group which uses regression model as basis for predicted rates

noperiods

A list of candidate number of periods in prediction base (e.g 4:6).If the goodness of fit test is rejected based on the widest base (e.g. 6 periods), the first period is exclude etc. Use a fixed number to force a specific prediction base. If e.g. noperiods = 5, predictions is based on the last 5 five-year periods, irrespective of the result a goodness of fit evaluation

recent

Project average trend or use the slope for the last 10 years? (If recent = FALSE, average trend for the whole observation period is used, if recent = TRUE, the slope from the last 10 years is used. If NULL (default) the choice is based on a significance test for departure from linear trend

cuttrend

Cut trend in predictions? Default is 0 %, 25 %, 50 %, 75 %, 75 % cut in drift (a vector of proportions of drift to cut in each projection period)

linkfunc

Link function to use in the model. Default is special version used in the Nordpred project ('power5'), where the link is g(x)=x^0.2, while the alternative is the poisson function ('poisson'), where the link is g(x)=log(x)

x

an object to test for class nordpred

Details

For details of the choice of prediction base, significance test for using recent slope, and for the power5 model, see Moller B., Fekjaer H. et al. (2002), see references.

Value

nordpred returns an object of class nordpred (see nordpred.object).

is.nordpred returns TRUE if input object is of class nordpred, FALSE otherwise.

Note for S-plus

Powerlink is made via a special modification in S-PLUS. This works fine for the point estimates, but the variance estimates found via the glm-objects are wrong. For variance estimates, we would rather recommend using R.

Author(s)

Harald Fekjaer and Bjorn Moller (Cancer Registry of Norway)

References

See Also

Other nordpred: is.nordpred.estimate, nordpred.estimate; nordpred.estimate.object; nordpred.getpred; nordpred.object; nordpred.prediction; plot.nordpred; print.nordpred.estimate; print.nordpred; summary.nordpred

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
# data (Colon cancer for Norwegian males)
indata

# Create dataset with observed and predicted population
inpop <- cbind(inpop1, inpop2)

# Fit model & predict new incidence:
res <- nordpred(indata, inpop, startestage = 5, startuseage = 6,
         cuttrend = c(0, .25, .5, .75, .75))
res2 <- nordpred(indata, inpop, startestage = 5, startuseage = 6,
     cuttrend = c(0, .25, .5, .75, .75), linkfunc = 'poisson')

# Print / get results:
print(res)
nordpred.getpred(res)
summary(res, printpred = FALSE)

# Get results with standardisation:
wstand <- c(0.12, 0.1, 0.09, 0.09, 0.08, 0.08, 0.06, 0.06, 0.06, 0.06,0.05,
            0.04, 0.04, 0.03, 0.02, 0.01, 0.005, 0.005)
round(nordpred.getpred(res, incidence = TRUE, standpop = NULL), 2)
round(nordpred.getpred(res, incidence = TRUE, standpop = wstand), 2)

# Plot results:
plot(res, standpop = wstand)

# Plot results with power5 and poisson links:
plot(res2, standpop = wstand)
plot(res, new = FALSE, lty = c(1,2), standpop = wstand)

# Make estimates:
est <- nordpred.estimate(cases = indata, pyr = inpop, noperiod = 4, startestage = 5)

# Different cut trend scenarios, using average drift (recent = FALSE):
plot(nordpred.prediction(est, startuseage = 6, cuttrend = c(0, 0, 0, 0, 0), recent = FALSE),
     standpop = wstand, new = TRUE)
plot(nordpred.prediction(est, startuseage = 6, cuttrend = c(1, 1, 1, 1, 1), recent = FALSE),
     standpop = wstand, new = FALSE, lty = c(1, 2))
plot(nordpred.prediction(est, startuseage = 6, cuttrend = c(0, .25, .5, .75, .75),
     recent = FALSE), standpop = wstand, new = FALSE, lty = c(1, 4))

haraldwf/nordpred documentation built on May 17, 2019, 2:41 p.m.