Figure 4
From: BioPhysConnectoR: Connecting Sequence Information and Biophysical Models
Timings. We tested the routine simc for breaking individual contacts and computing the Frobenius norm of the covariance matrix with respect to the original matrix using different numbers of cores. We compared the speed-up of the parallelization provided by the R packages multicore [31] and snow[18]. The elapsed time for both is fitted to a scaling law of the form tCPU ≈ a + c (#cores)-1 with some unimportant constants a, c and tCPU the total CPU time, and #cores the number of multi-cores used.