※ 引述《brooky (未够班)》之铭言：
: ※ 引述《Yyh (我爱批踢踢)》之铭言：
: : I don't know why you trace those two function in grow.c @@?
: : I am going to trace "BDRgrow1",
: I trace these two functions because I think BDRgrow is just allocating
: some memory and initialize some variable.
: I think "divide_node(0)" , which is called in BDRgrow,
: might contain more information.
: "split_disc" and "split_cont" were called in "divide_node"
: So I start trying to understand these two functions.
: : because it appears in grow.q as "fit <- .C("BDRgrow1"......"
: : bit it confused me too >"<
: : Anyone have better tracing skill?
: I have put "tree" aside and decided to trace rpart first.
: "rpart" seems easier than "tree" because it have more comments.
did anyone know what is "Complexity Parameter " in rpart, this seems an
important info