Rajesh Ilango

At the moment we have no such plans. We will keep this in mind and develop such scripts in Python instead.

Thanks. Working on a patch and testing.

Sorry for the delayed response. I will trying to reproduce this issue. I will get back to you soon.

If you are working from source, please follow these steps. Terminal 1 ---------- 1. ./launch.sh start This will download pre-reqs and start the application. Terminal 2 ---------- Default IP address...

@Intron7 Please provide us the OS version used in both case. We noticed **"OS call failed or operation not supported on this OS"** in the error message. RMM is supported...

I can reproduce this issue. Additional stacktrace info ipp1-0129:2216 :0:2216] Caught signal 11 (Segmentation fault: address not mapped to object at address 0x10) ==== backtrace (tid: 2216) ==== 0 /root/conda/rapids-0.18/lib/python3.8/site-packages/ucp/_libs/../../../../libucs.so.0(ucs_handle_error+0x115)...

@Intron7 Please try with USE_FIRST_N_CELLS = 700000 Meanwhile, we will be creating a bug with RAPIDS regarding this issue.

Recently a multi-GPU version of the notebook was added to the examples. https://github.com/clara-parabricks/rapids-single-cell-examples/blob/master/notebooks/1M_brain_gpu_analysis_multigpu.ipynb This allows much larger datasets. Please check if this helps.

@m0nib Sorry for the delay in response. Can you please send us the error you are getting from adata.obsm['X_tsne'] = TSNE().fit_transform(adata.X[:,:tsne_n_pcs])?

Hi Jegathesan, Please point us to the cell in the notebook where this issue appears. Regards, Rajesh K Ilango