ValueError: cannot reshape array of size 0 into shape (0,newaxis)

[Yaoyx]

Hello,

I tried to print the snippet, and it's an empty array. I appreciate for any suggestions.

Thank you,
Yao

[Yaoyx]

hit_scorer = densityadapted_hitscoring.MakeHitScorer(
    patterns=orig_patterns,
    target_seqlet_size=11,
    bg_freq=np.mean(test_inputs[target_ind][None,:,:].astype('int'), axis=(0,1)),
    task_names_and_signs=[("task0",1)],
    n_cores=10)

I found decreasing target_seqlet_size to 11 helped, and here's my setting for getting tfmodisco_results:

tfmodisco_results = modisco.tfmodisco_workflow.workflow.TfModiscoWorkflow(
                    sliding_window_size=18,
                    flank_size=8,
                    max_seqlets_per_metacluster=40000,
                    seqlets_to_patterns_factory=
                     modisco.tfmodisco_workflow.seqlets_to_patterns.TfModiscoSeqletsToPatternsFactory(
                        trim_to_window_size=24,
                        initial_flank_to_add=8,
                        n_cores=10)
                )

I am a bit confused about the relationship between parameters in these two functions. It seems that target_seqlet_size in densityadapted_hitscoring.MakeHitScorer should be smaller than trim_to_window_size in modisco.tfmodisco_workflow.workflow.TfModiscoWorkflow, but I am not sure about the ratio between them.