Please help with ：IndexError: tuple index out of range for cell_states_to_model
#119
by
jinbo1129
- opened
hello, thanks for this great tool and your quick response !!!
I try to do the in_silico_perturbation with my own dataset, where are only two cellstates: normal and healthy, so I edit the code to:
# the embedding towards non-failing (nf) state
isp = InSilicoPerturber(perturb_type="delete",
perturb_rank_shift=None,
genes_to_perturb="all",
combos=0,
anchor_gene=None,
model_type="CellClassifier",
num_classes=2, ## change according to my own data !!!
emb_mode="cell",
cell_emb_style="mean_pool",
cell_states_to_model={"disease":(["Normal"],["Disease"])},
max_ncells=2000,
emb_layer=0,
forward_batch_size=400,
nproc=16)
Than I met with the errors:
IndexError Traceback (most recent call last)
551 # get dictionary of average cell state embeddings for comparison
--> 552 state_embs_dict = get_cell_state_avg_embs(model,
553 filtered_input_data,
554 self.cell_states_to_model,
555 layer_to_quant,
556 self.gene_token_dict,
557 self.forward_batch_size,
558 self.nproc)
559 self.in_silico_perturb(model,
560 filtered_input_data,
561 layer_to_quant,
562 state_embs_dict,
563 output_directory,
564 output_prefix)
--> 163 possible_states = [value[0]+value[1]+value[2] for value in cell_states_to_model.values()][0]
164 state_embs_dict = dict()
165 for possible_state in possible_states:
IndexError: tuple index out of range
Should I change the codes of "in_silico_perturber.py" according to my own dataset ?
Thanks !!!
cell_states_to_model: None, dict
Cell states to model if testing perturbations that achieve goal state change.
Single-item dictionary with key being cell attribute (e.g. "disease").
Value is tuple of three lists indicating start state, goal end state, and alternate possible end states.
If no alternate possible end states, third list should be empty (i.e. the third list should be []).
jinbo1129
changed discussion status to
closed