Plugin error from feature-classifier - do I need more RAM?

Hi, I am new! Sorry if this question is silly... thanks in advance!
I got this error when I tried to run a silva V3V4 classifier:

Plugin error from feature-classifier: Unable to allocate 2.77 GiB for an array with shape (372064256,) and data type float64 Debug info has been saved to /tmp/qiime2-q2cli-err-fyei80v1.log

Does this means my computer's RAM is not enough for the run? My computer has only 8GB RAM in total so I can only allocate 5GB for its usage.
I am using QIIME 2 2020.2 in Virtualbox.

The command I ran:

qiime feature-classifier classify-sklearn \
> --i-classifier /media/sf_Shared_Folders/silva138_AB_V3-V4_classifier.qza \
> --i-reads /media/sf_Shared_Folders/rep-seqs.qza \
> --o-classification /media/sf_Shared_Folders/taxonomy.qza

The log:

Traceback (most recent call last):
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/q
2cli/", line 339, in call
results = action(**arguments)
File "", line 2, in classify_sklearn
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/q
iime2/sdk/", line 234, in bound_callable
callable_args[name] = artifact._view(
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/q
iime2/sdk/", line 331, in _view
result = transformation(self._archiver.data_dir)
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/qiime2/core/", line 70, in transformation
new_view = transformer(view)
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/q2_feature_classifier/", line 72, in _1
pipeline = joblib.load(os.path.join(dirname, 'sklearn_pipeline.pkl'))
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/joblib/", line 587, in load
obj = _unpickle(fobj, filename, mmap_mode)
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/joblib/", line 506, in _unpickle
obj = unpickler.load()
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/", line 1212, in load
File "/home/qiime2/miniconda/envs/qiime2-2022.2/lib/python3.8/site-packages/j:

1 Like

Hi @hewsy ,

Correct! You will probably need at least 8 GB of RAM, but probably more like 12-16GB as you are working with SILVA. If you cannot access a more powerful computer, you might need to use a different classifier (e.g., greengenes or NCBI refseqs, which are smaller and hence take less RAM)

Good luck!

1 Like

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.