Hello,
I was trying to run the taxonomy using the pre-trained classifier.
qiime feature-classifier classify-sklearn
–i-reads dada2-demux-representative-seqs.qza
–i-classifier silva-138-99-nb-classifier.qza
–o-classification cement-taxonomy
I got the following error
Plugin error from feature-classifier:
Unable to allocate 4.51 GiB for an array with shape (604749824,) and data type float64
Debug info has been saved to /tmp/qiime2-q2cli-err-7l45p5dl.log
@sakshi, I'm pretty sure this is an out of memory error, indicating that your system has insufficient RAM available. Silva is pretty memory- and storage-intensive, so this is a common problem.
If you're working in a virtual machine, try increasing the amount of RAM available to the VM.
If you're running this command on a supercompute cluster (or similar), try asking the scheduler for more RAM
If you have access to an institutional cluster but aren't using it, try running this command there
If you don't, but your personal computer doesn't have enough RAM, you could rent a server and natively install QIIME 2 on that server. Ubuntu servers are pretty straightforward to set up, and this is usually surprisingly cheap.
If you don't need silva for your study, you could use a smaller/lighter classifier (like greengenes)
If none of the above work for you, you could ask a friend with more RAM to run it for you overnight/over the weekend.
If your friends are all on their computers 24/7, installing more RAM will probably fix the issue. How much RAM will do? I suspect that depends on the size of your data set.
