multinma::run() to return the knot locations pretty please :)

[chumbleycode]

It would be super helpful for nma() to return the knot locations that it uses in any spline-based likelihood! This would help with sensistivity analysis and QC.

[dmphillippo]

Hi @chumbleycode, these are already available - if fit is your fitted model object output from nma() then the basis objects are stored in a named list (one for each study in the network) in fit$basis. The bases are created using splines2::Mspline(), and you can access the knots like attr(fit$basis$study1, "knots") or splines2::knots(fit$basis$study1).

You can also use the make_knots() function to check the knot locations before having fitted the model.

[chumbleycode]

Thank you!