hpdbscan
hpdbscan copied to clipboard
Memory issue
Great work, thanks.
I have an issue about the memory. In the example, the features have small dimension, eg, 3. When I try larger dimension, I will face out of memory issue. e.g., I tried data = np.random.rand(100, 300) it takes more than 100G memory.
Do you have any idea about this issue ? Thanks.
Hi @daodaofr,
I am aware of this issue. It has something to do with the way high numbers of dimensions are handled internally. Consider it a bug for now. I am looking into it. For now, however, it is simply not possible to process this or you have to buy a machine with a lot of RAM :P.
Cheers