print.topmod {BMS} | R Documentation |
Print method for objects of class 'topmod', typically the best models stored in a 'bma' object
## S3 method for class 'topmod' print(x, ...)
x |
an object of class 'topmod' - see |
... |
additional arguments passed to |
See pmp.bma
for an explanation of likelihood vs. MCMC frequency concepts
if x
contains more than one model, then the function returns a 2-column matrix:
Row Names |
show the model binaries in hexcode (cf. |
Column 'Marg.Log.Lik' |
shows the marginal log-likelihoods of the models in |
Column 'MCMC Freq' |
shows the MCMC frequencies of the models in |
if x
contains only one model, then more detailed information is shown for this model:
first line |
'Model Index' provides the model binary in hexcode, 'Marg.Log.Lik' its marginal log likelhood, 'Sampled Freq.' how often it was accepted (function |
Estimates |
first column: covariate indices included in the model, second column: posterior expected value of the coefficients, third column: their posterior standard deviations (excluded if no coefficients were stored in the topmod object - cf. argument |
Included Covariates |
the model binary |
Additional Statistics |
any custom additional statistics saved with the model |
Martin Feldkircher and Stefan Zeugner
topmod
for creating topmod objects, bms
for their typical use, pmp.bma
for comparing posterior model probabilities
Check http://bms.zeugner.eu for additional help.
# do some small-scale BMA for demonstration data(datafls) mm=bms(datafls[,1:10],nmodel=20) #print info on the best 20 models print(mm$topmod) print(mm$topmod,digits=10) #equivalent: cbind(mm$topmod$lik(),mm$topmod$ncount()) #now print info only for the second-best model: print(mm$topmod[2]) #compare 'Included Covariates' to: topmodels.bma(mm[2]) #and to as.vector(mm$topmod[2]$bool_binary())