Fit a Stochastic Mortality Model
Fit a Stochastic Mortality Model to a given data set. The fitting is done
using package gnm.
## S3 method for class 'StMoMo' fit(object, data = NULL, Dxt = NULL, Ext = NULL, ages = NULL, years = NULL, ages.fit = NULL, years.fit = NULL, oxt = NULL, wxt = NULL, start.ax = NULL, start.bx = NULL, start.kt = NULL, start.b0x = NULL, start.gc = NULL, verbose = TRUE, ...)
object |
an object of class |
data |
an optional object of type StMoMoData containing information on
deaths and exposures to be used for fitting the model. This is typically created
with function |
Dxt |
optional matrix of deaths data. |
Ext |
optional matrix of observed exposures of the same dimension of
|
ages |
optional vector of ages corresponding to rows of |
years |
optional vector of years corresponding to rows of |
ages.fit |
optional vector of ages to include in the fit. Must be a
subset of |
years.fit |
optional vector of years to include in the fit. Must be a
subset of |
oxt |
optional matrix/vector or scalar of known offset to be used in fitting the model. This can be used to specify any a priori known component to be added to the predictor during fitting. |
wxt |
optional matrix of 0-1 weights to be used in the fitting process.
This can be used, for instance, to zero weight some cohorts in the data.
See |
start.ax |
optional vector with starting values for α_x. |
start.bx |
optional matrix with starting values for β_x^{(i)}. |
start.kt |
optional matrix with starting values for κ_t^{(i)}. |
start.b0x |
optional vector with starting values for β_x^{(0)}. |
start.gc |
optional vector with starting values for γ_c. |
verbose |
a logical value. If |
... |
arguments to be passed to or from other methods. This can be
used to control the fitting parameters of |
Fitting is done using function gnm within package
gnm. This is equivalent to minimising (maximising) the deviance
(log-likelihood) of the model. Ages and years in the data should be of
type numeric. Data points with zero exposure are assigned a zero weight
and are ignored in the fitting process. Similarly, NA are assigned a
zero weight and ignored in the fitting process. Parameter estimates can be
plotted using function plot.fitStMoMo.
A list with class "fitStMoMo" with components:
model |
the object of class |
ax |
vector with the fitted values of the static age function
α_x. If the model does not have a static age function or
failed to fit this is set to |
bx |
matrix with the values of the period age-modulating functions
β_x^{(i)}, i=1, ..., N. If the i-th age-modulating
function is non-parametric (e.g. as in the Lee-Carter model)
|
kt |
matrix with the values of the fitted period indexes
κ_t^{(i)}, i=1, ..., N. |
b0x |
vector with the values of the cohort age-modulating function
β_x^{(0)}. If the age-modulating function is non-parametric
|
gc |
vector with the fitted cohort index γ_{c}.
If the model does not have a cohort effect or failed to fit this is set
to |
data |
StMoMoData object provided for fitting the model. |
Dxt |
matrix of deaths used in the fitting. |
Ext |
matrix of exposures used in the fitting. |
oxt |
matrix of known offset values used in the fitting. |
wxt |
matrix of 0-1 weights used in the fitting. |
ages |
vector of ages used in the fitting. |
years |
vector of years used in the fitting. |
cohorts |
vector of cohorts used in the fitting. |
fittingModel |
output from the call to |
loglik |
log-likelihood of the model. If the fitting failed to
converge this is set to |
deviance |
deviance of the model. If the fitting failed to
converge this is set to |
npar |
effective number of parameters in the model. If the fitting
failed to converge this is set to |
nobs |
number of observations in the model fit. If the fitting
failed to converge this is set to |
fail |
|
conv |
|
@seealso StMoMoData, genWeightMat,
plot.fitStMoMo, EWMaleData
# Lee-Carter model only for older ages
LCfit <- fit(lc(), data = EWMaleData, ages.fit = 55:89)
plot(LCfit)
# Use arguments Dxt, Ext, ages, years to pass fitting data
LCfit <- fit(lc(), Dxt = EWMaleData$Dxt, Ext = EWMaleData$Ext,
ages = EWMaleData$ages, years = EWMaleData$years,
ages.fit = 55:89)
plot(LCfit)
# APC model weigthing out the 3 first and last cohorts
wxt <- genWeightMat(EWMaleData$ages, EWMaleData$years, clip = 3)
APCfit <- fit(apc(), data = EWMaleData, wxt = wxt)
plot(APCfit, parametricbx = FALSE, nCol = 3)
# Set verbose = FALSE for silent fitting
APCfit <- fit(apc(), data = EWMaleData, wxt = wxt,
verbose = FALSE)
## Not run:
# Poisson Lee-Carter model with the static age function set to
# the mean over time of the log-death rates
constLCfix_ax <- function(ax, bx, kt, b0x, gc, wxt, ages){
c1 <- sum(bx, na.rm = TRUE)
bx <- bx / c1
kt <- kt * c1
list(ax = ax, bx = bx, kt = kt, b0x = b0x, gc = gc)
}
LCfix_ax <- StMoMo(link = "log", staticAgeFun = FALSE,
periodAgeFun = "NP", constFun = constLCfix_ax)
LCfix_axfit <- fit(LCfix_ax, data= EWMaleData,
oxt = rowMeans(log(EWMaleData$Dxt / EWMaleData$Ext)))
plot(LCfix_axfit)
## End(Not run)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.