But I kept receiving error messages saying 'Error in quadprog::solve.QP(Dmat, t(Xmat[obs, ]) %*% Y[obs, i], Amat, : matrix D in quadratic function is not positive definite!'.
Or, when I use 'data<-2^exprs(gsm[[1]])', I received 'Error in quadprog::solve.QP(Dmat, t(Xmat[obs, ]) %*% Y[obs, i], Amat, : constraints are inconsistent, no solution!'
I'm not sure if I apply TOAST correctly, or if you have done some normalization or standardization beforehand.
Hi Ziyi,
Greeting! I have encountered some problems when trying to replicate the result of applying TOAST on GSE19830 in your paper. A glimpse what I ran in R:
library(GEOquery) library(TOAST)
gsm <- getGEO("GSE19830") data<-exprs(gsm[[1]]) K=3 set.seed(1234) outRF1 <- csDeconv(data, K, TotalIter = 100, bound_negative = TRUE)
But I kept receiving error messages saying 'Error in quadprog::solve.QP(Dmat, t(Xmat[obs, ]) %*% Y[obs, i], Amat, : matrix D in quadratic function is not positive definite!'.
Or, when I use 'data<-2^exprs(gsm[[1]])', I received 'Error in quadprog::solve.QP(Dmat, t(Xmat[obs, ]) %*% Y[obs, i], Amat, : constraints are inconsistent, no solution!'
I'm not sure if I apply TOAST correctly, or if you have done some normalization or standardization beforehand.
Thanks for your consideration!
Best,
Niccolo