Finite mixture modelling of geochronological datasets
Implements the discrete mixture modelling algorithms of Galbraith and Laslett (1993) and applies them to fission track and other geochronological datasets.
peakfit(x, ...) ## Default S3 method: peakfit(x, k = "auto", sigdig = 2, log = TRUE, alpha = 0.05, ...) ## S3 method for class 'fissiontracks' peakfit(x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, alpha = 0.05, ...) ## S3 method for class 'UPb' peakfit( x, k = 1, type = 4, cutoff.76 = 1100, cutoff.disc = discfilter(), common.Pb = 0, exterr = TRUE, sigdig = 2, log = TRUE, alpha = 0.05, ... ) ## S3 method for class 'PbPb' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, common.Pb = 0, alpha = 0.05, ... ) ## S3 method for class 'ArAr' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, i2i = FALSE, alpha = 0.05, ... ) ## S3 method for class 'ThPb' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, i2i = FALSE, alpha = 0.05, ... ) ## S3 method for class 'KCa' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, i2i = FALSE, alpha = 0.05, ... ) ## S3 method for class 'ReOs' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, i2i = TRUE, alpha = 0.05, ... ) ## S3 method for class 'SmNd' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, i2i = TRUE, alpha = 0.05, ... ) ## S3 method for class 'RbSr' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, i2i = TRUE, alpha = 0.05, ... ) ## S3 method for class 'LuHf' peakfit( x, k = 1, exterr = TRUE, sigdig = 2, log = TRUE, i2i = TRUE, alpha = 0.05, ... ) ## S3 method for class 'ThU' peakfit( x, k = 1, exterr = FALSE, sigdig = 2, log = TRUE, i2i = TRUE, alpha = 0.05, detritus = 0, ... ) ## S3 method for class 'UThHe' peakfit(x, k = 1, sigdig = 2, log = TRUE, alpha = 0.05, ...)
x |
either an |
... |
optional arguments (not used) |
k |
the number of discrete age components to be
sought. Setting this parameter to |
sigdig |
number of significant digits to be used for any legend in which the peak fitting results are to be displayed. |
log |
take the logs of the data before applying the mixture model? |
alpha |
cutoff value for confidence intervals |
exterr |
propagate the external sources of uncertainty into the component age errors? |
type |
scalar indicating whether to plot the
^{207}Pb/^{235}U age ( |
cutoff.76 |
the age (in Ma) below which the
^{206}Pb/^{238}U and above which the
^{207}Pb/^{206}Pb age is used. This parameter is
only used if |
cutoff.disc |
discordance cutoff filter. This is an object of
class |
common.Pb |
common lead correction:
|
i2i |
‘isochron to intercept’: calculates the initial (aka
‘inherited’, ‘excess’, or ‘common’)
^{40}Ar/^{36}Ar, ^{40}Ca/^{44}Ca,
^{207}Pb/^{204}Pb, ^{87}Sr/^{86}Sr,
^{143}Nd/^{144}Nd, ^{187}Os/^{188}Os,
^{230}Th/^{232}Th, ^{176}Hf/^{177}Hf or
^{204}Pb/^{208}Pb ratio from an isochron
fit. Setting |
detritus |
detrital ^{230}Th correction (only applicable
when
|
Consider a dataset of n dates \{t_1, t_2, ..., t_n\}
with analytical uncertainties \{s[t_1], s[t_2], ...,
s[t_n]\}. Define z_i = \log(t_i) and s[z_i] =
s[t_i]/t_i. Suppose that these n values are derived from a
mixture of k>2 populations with means
\{μ_1,...,μ_k\}. Such a discrete mixture may be
mathematically described by P(z_i|μ,ω) = ∑_{j=1}^k
π_j N(z_i | μ_j, s[z_j]^2 ) where π_j is the
proportion of the population that belongs to the j^{th}
component, and π_k=1-∑_{j=1}^{k-1}π_j. This equation
can be solved by the method of maximum likelihood (Galbraith and
Laslett, 1993). IsoplotR implements the Bayes Information
Criterion (BIC) as a means of automatically choosing k. This
option should be used with caution, as the number of peaks steadily
rises with sample size (n). If one is mainly interested in
the youngest age component, then it is more productive to use an
alternative parameterisation, in which all grains are assumed to
come from one of two components, whereby the first component is a
single discrete age peak (\exp(m), say) and the second
component is a continuous distribution (as descibed by the
central age model), but truncated at this discrete
value (Van der Touw et al., 1997).
Returns a list with the following items:
a 3 x k matrix with the following rows:
t: the ages of the k peaks
s[t]: the estimated uncertainties of t
ci[t]: the widths of approximate 100(1-α)\%
confidence intervals for t
a 2 x k matrix with the following rows:
p: the proportions of the k peaks
s[p]: the estimated uncertainties (standard errors) of
p
the log-likelihood of the fit
a vector of text expressions to be used in a figure legend
Galbraith, R.F. and Laslett, G.M., 1993. Statistical models for mixed fission track ages. Nuclear Tracks and Radiation Measurements, 21(4), pp.459-470.
van der Touw, J., Galbraith, R., and Laslett, G. A logistic truncated normal mixture model for overdispersed binomial data. Journal of Statistical Computation and Simulation, 59(4):349-373, 1997.
data(examples) peakfit(examples$FT1,k=2) peakfit(examples$LudwigMixture,k='min')
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.