I am at the last stage of publishing my package. Since I did the initial submission to CRAN, they commented on the verification of examples:
Moreover,
* checking examples ... [169s/169s] OK
Examples with CPU or elapsed time > 5s
user system elapsed
function1 66.240 0.004 66.248
function2 54.404 0.020 54.430
function3 47.060 0.192 47.252
and the CRAN policy asks just a few seconds: can you reduce the above runtime?
Each function listed above consists of an evaluator using an iterative process, and they use fairly intensive calculations.
Here's the deal: I can limit, for illustrative example only, the number of iterations with the parameter of these functions, but this would provide estimates that did not converge. In addition, the illustrative data set used for the evaluation process is quite small, so I cannot get too much better without limiting the number of iterations.
, ... / ? , ?
. .