Hello All, I am attempting to perform cross-validation to test my BGPCA and any misclassification rates. The coding examples online are based on general linear models, so of course I need to modify the code for my shape data. I've tried using PCs and eigenscores in several ways in my coding and have gotten nothing but error messages. I understand that I need to write the code so it writes response~predictor, but am obviously not getting how to manipulate that with my data. Any insight would be much appreciated.
ALL.means = superimposed configurations ALL.means.shape = shape data ALL.means.super = superimpose consensus configuration mPCAALL$x = PCs E = eigenscores > glmFit <- glm(ALL.means ~ mPCAALL$x[,1:2], data=NULL) Error in x[good, , drop = FALSE] : (subscript) logical subscript too long > glmFit <- glm(ALL.means.shape ~ mPCAALL$x[,1:2], data=NULL) Error in model.frame.default(formula = ALL.means.shape ~ mPCAALL$x[, 1:2], : variable lengths differ (found for 'mPCAALL$x[, 1:2]') > glmFit <- glm(mPCAALL$x[,1:2] ~ log(mcsize), data=NULL) Error in x[good, , drop = FALSE] : (subscript) logical subscript too long > glmFit <- glm(ALL.means ~ E, data=NULL) Error in x[good, , drop = FALSE] : (subscript) logical subscript too long > glmFit <- glm(ALL.means.shape ~ E, data=NULL) Error in model.frame.default(formula = ALL.means.shape ~ E, data = NULL, : variable lengths differ (found for 'E') > glmFit <- glm(mPCAALL$x[,1:2] ~ E[,1:2], data=NULL) Error in x[good, , drop = FALSE] : (subscript) logical subscript too long > glmFit <- glm(mPCAALL$x ~ E, data=NULL) Error in x[good, , drop = FALSE] : (subscript) logical subscript too long > glmFit <- glm(ALL.means ~ mPCAALL$x, data=NULL) Error in x[good, , drop = FALSE] : (subscript) logical subscript too long > glmFit <- glm(ALL.means.shape ~ mPCAALL$x, data=NULL) Error in model.frame.default(formula = ALL.means.shape ~ mPCAALL$x, data = NULL, : variable lengths differ (found for 'mPCAALL$x') > glmFit <- glm(ALL.means.super ~ mPCAALL$x, data=NULL) Error in model.frame.default(formula = ALL.means.super ~ mPCAALL$x, data = NULL, : invalid type (list) for variable 'ALL.means.super' Thank you, Brenna Hays Research Assistant, M.S. Student Nova Southeastern University, Oceanographic Center -- MORPHMET may be accessed via its webpage at http://www.morphometrics.org --- You received this message because you are subscribed to the Google Groups "MORPHMET" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
