Memory error in "select 2D classes"

I’m supporting a user that is getting a memory error when running “select 2D classes”. I found 2 topics that address this problem from back in Feb but the patch you made in version 2.15.0 (which we are running) doesn’t seem to fix this. I’ve appended the traceback. The job runs on the master node which it looks like it should. Any fix/work-around?

[CPU: 34.14 GB] Traceback (most recent call last): File “cryosparc2_master/cryosparc2_compute/”, line 89, in File “cryosparc2_compute/jobs/”, line 884, in passthrough_outputs dset = load_input_group(input_group_name, passthrough_result_names, allow_passthrough=True) File “cryosparc2_compute/jobs/”, line 553, in load_input_group dsets = [load_input_connection_slots(input_group_name, keep_slot_names, idx, allow_passthrough=allow_passthrough) for idx in range(num_connections)] File “cryosparc2_compute/jobs/”, line 524, in load_input_connection_slots dsets = [load_input_connection_single_slot(input_group_name, slot_name, connection_idx, allow_passthrough=allow_passthrough) for slot_name in slot_names] File “cryosparc2_compute/jobs/”, line 516, in load_input_connection_single_slot d = load_output_result_dset(_project_uid, output_result, slotconnection[‘version’], slot_name) File “cryosparc2_compute/jobs/”, line 469, in load_output_result_dset d.from_file(abspath) File “cryosparc2_compute/”, line 543, in from_file = Dataset._convert_strings(indata, to_fixed=False) File “cryosparc2_compute/”, line 597, in _convert_strings newdata = Dataset.allocate(len(data), new_dtype_desc) File “cryosparc2_compute/”, line 125, in allocate ) for t in dtype_descr MemoryError

HI @GeneF,

How many particles was this user trying to load into this Select2D job?


Thanks for the reply.

The researcher is trying to load about 35 million binned particles.

The server had about 180GB of memory free when it tried to do the classification.