内核PCA与Kernlab r，内存不足错误

Question

尝试在12 * 250,000数据集的r中应用带有Kernlab的内核PCA

 kpc <- kpca(~.,csv.data,kernel="rbfdot",kpar=list(sigma=0.2),features=2)

收到波纹管错误

Error: cannot allocate vector of size 604.1 Gb
In addition: Warning messages:
1: In crossprod(t(x)) :
  Reached total allocation of 16305Mb: see help(memory.size)
2: In crossprod(t(x)) :
  Reached total allocation of 16305Mb: see help(memory.size)
3: In crossprod(t(x)) :
  Reached total allocation of 16305Mb: see help(memory.size)
4: In crossprod(t(x)) :
  Reached total allocation of 16305Mb: see help(memory.size)

有任何建议如何设法在如此大的数据集上运行此算法？

感谢。