sim_par {mina} | R Documentation |
Function for community similarity calculation used by 'tina', modified from https://github.com/defleury/Schmidt_et_al_2016_community_similarity/blob/ master/functions.community_similarity.R
sim_par(x, y, sim_method = "w_ja", threads = 80, nblocks = 400)
x |
An quantitative matrix. |
y |
The Cij matrix, which is correlation matrix of adjusted sparcc matrix of x. |
sim_method |
The method for similarity, "w_ja" and "uw_ja" are available for weighted and unweighted Jaccard similarity respectively. |
threads |
The number of threads used for parallel running, 80 by default. |
nblocks |
The number of row / column for splitted sub-matrix, 400 by default. |
s The output similarity matrix.
## Not run: data(maize_asv) maize_tab <- maize_asv[1 : 1000, 1 : 200] asv <- norm_tab(maize_tab, method = "raref", depth = 100) asv[is.na(asv)] <- 0 asv_sparcc <- sparcc(asv, threads = 8, nblocks = 40) tmp.S <- adj(asv_sparcc, method = "spearman") y <- 0.5 * (tmp.S + 1) s <- sim_par(asv_sparcc, y, sim_method = "w_ja", threads = 8, nblocks = 40) ## End(Not run)