Converts parameter values between two different parameterisations (described in Details below) of the linked stress release model.
linksrm_convert(params, abc=TRUE)a vector of parameter values of length \(n^2+2n\), where \(n\) is the number of regions in the model.
logical. If TRUE (default), then the input value of params is that of the abc parameterisation. See Details for further explanation.
A list object with the following components is returned:
vector as specified in the function call.
vector of length \(n\) as in the abc parameterisation.
vector of length \(n\) as in the abc parameterisation.
n by \(n\) matrix as in the abc parameterisation.
vector of length \(n\) as in the alternative parameterisation.
vector of length \(n\) as in the alternative parameterisation.
vector of length \(n\) as in the alternative parameterisation.
n by \(n\) matrix with ones on the diagonal as in the alternative parameterisation.
If abc == TRUE, the conditional intensity for the \(i\)th region is assumed to have the form
$$
\lambda_g(t,i | {\cal H}_t) = \exp\left\{ a_i + b_i\left[t - \sum_{j=1}^n c_{ij} S_j(t)\right]\right\}
$$
with params\( = (a_1, \cdots, a_n, b_1, \cdots, b_n, c_{11}, c_{12}, c_{13}, \cdots, c_{nn})\).
If abc == FALSE, the conditional intensity for the \(i\)th region is assumed to have the form
$$
\lambda_g(t,i | {\cal H}_t) = \exp\left\{ \alpha_i + \nu_i\left[\rho_i t - \sum_{j=1}^n \theta_{ij} S_j(t)\right]\right\}
$$
where \(\theta_{ii}=1\) for all \(i\), \(n = \sqrt{\code{length(params)} + 1} - 1\), and
params$$ = (\alpha_1, \cdots, \alpha_n, \nu_1, \cdots, \nu_n, \rho_1, \cdots, \rho_n, \theta_{12}, \theta_{13}, \cdots, \theta_{1n}, \theta_{21}, \theta_{23}, \cdots, \theta_{n,n-1}).$$