pro hr_blk_intrp,hrb,logtn,logln,teffi,logli,lnprob,bias=bias ; This routine accepts the HR diagram probability block data hrb, defined on a ; grid with coordinates logtn, logln ; (typically obtained from a file written by mk_hrblock.pro ; and two vectors of length npt of parameters ; teffi, logli, for which probability densities are desired. ; It interpolates into ; the hrb table to generate the ln of the corresponding probabilities ; lnprob(npt) ; If keyword bias is set, its value multiplies an additive linear ; function that (if bias is positive) makes giants relatively more ; probable than dwarfs, a priori. Reasonable values of bias (which do not ; distort the a priori probabilities in a crazy way) are roughly 1 to 10. ; constants tsun=5777. bt=-0.992 ; coeff of logT for bias function bl=0.128 ; coeff of logL for bias function ; find indices into the HR block for the desired inputs sz=size(hrb) nt=sz(1) nl=sz(2) iit=findgen(nt) iil=findgen(nl) npt=n_elements(teffi) lteffi=alog10(teffi/tsun) jt=interpol(iit,logtn,lteffi) jl=interpol(iil,logln,logli) lnprob=fltarr(npt) lnprob=interpolate(hrb,jt,jl,/cubic) if(keyword_set(bias)) then lnprob=lnprob+bias*(bt*lteffi+bl*logli) end