Dear R Users, R Core Team,
I currently wonder how to predict the probability of an event with new data
resulting from a finite mixture.
I read the documentation of the flexmix package and the examples of applications
provided on CRAN but I could not find how to predict (except
"manually" but I am looking for a simpler solution) the final
probability of the mixture (for each individual) with new data (I mean data
different from the ones I used to build the model).
I should have missed something but basically, I am fitting a 2-components
mixture model with logistic weights and logistic components (and different
explanatory variables, no identifiability problem). The flexmix object is then
used in predict() function with 'newdata' in argument, and the
predictions with these new data are obviously different depending on the
component which is assigned to new observations.
My question is: how can I access the information on clustering the new
observations (if I understood well, predictions are the predictions of the event
probability for each individual and each component, but these are not
predictions for cluster assignment...) ?
Thank you very much in advance for your answers,
Sincerely,
Xavier M.
Une messagerie gratuite, garantie à vie et des services en plus, ça vous tente ?
Je crée ma boîte mail www.laposte.net
[[alternative HTML version deleted]]
