createParams.Rd
Create parameters list for eNODAL_obj object.
createParams( gam_k = 29, gam = TRUE, test_method = "F", h_adj = TRUE, adj_method = "BH", sig_level = c(LC = 0.01, Linear = 0.01, Sig = 0.01, Interaction = 0.05, Meta1 = 0.05, Meta2 = 0.05), adapt = FALSE, test_func = "gam", sig_test = "LC-test", save_dist = FALSE, q0 = 0, knn.k = 40, dist_thresh = 0.4, eps = NULL, minPts = NULL, minClusterSize = NULL, kcluster = NULL, consensus_param = list(clmethod = "louvian", knn.k = 15, dist_thresh = 0.75), ... )
gam_k | Parameter k in GAM model. By default is 29. |
---|---|
gam | Whether use create GAM formula. By default is TRUE. If no continuous variable, set FALSE. |
test_method | Testing method. Can be chosen from "F", "globaltest", "Tmax", "Chisq", "Cp". By default is "F" |
h_adj | Indicator of whether use hierarchical adjustment. By default is TRUE |
adj_method | Pvalue adjust method. See p.adjust. By default is "BH". |
sig_level | Significance level of different test. By default is set below: LC = 0.01, Linear = 0.01, Interaction = 0.05, Meta1 = 0.05, Meta2 = 0.05. |
adapt | Indicator of whether to use adaptive way to decide testing type. |
test_func | Testing function used. By default is "gam". |
sig_test | Method for test between sig vs. non-sig. Can be choosen from "LC-test", "lm" or "gam. |
save_dist | Save distance result. By default is FALSE. |
q0 | Parameter q0 in apcluster. Byt default is 0. If it is NULL, will not use apcluster. |
knn.k | Number of k-nearest neighbour to create knn graph in Louvian algorithm. By default is 40, If it is NULL, will not use Louvian. |
dist_thresh | Threshold of distance when creating knn graph in Louvian algorithm. By default is 0.4, If it is NULL, will not use Louvian. |
eps | Parameter eps in dbscan. If it is NULL, will not use dbscan. |
minPts | Parameter minPts in dbscan. If it is NULL, will not use dbscan. |
minClusterSize | Parameter minClusterSize in dbscan. If it is NULL, will not use dbscan. |
kcluster | Number of clusters if using fixed number of clustering method. |
consensus_param | Parameter list for consensus clustering. Format: list(clmethod = ..., params = ...), available method and parameters listed above. By default is Louvian, with knn.k = 15, dist_thresh = 0.3. |
... | Other parameters can be passed to eNODAL_obj. |
A list contain parameters in eNODAL_obj object.