Hello Team,
Could you please provide the logs for the submissions 9761248 and 9761247.
Thanks in advance.
Regards,
Vijay N
Created by vijay kumar narsapuram narsapvi Hello @narsapvi,
Looking at the detailed trace, it seems your model is expecting different var_names (Ensembl IDs) than are in the training, validation, and final round datasets.
I've pasted an example of the AnnData.var field below. Our AnnData.var.index is gene\_name except when there were duplicates. In those cases we append the Ensembl ID onto the gene\_name (e.g. "TBHE ENSGXXXX").
```
gene_ids feature_types genome
index
MIR1302-2HG ENSG00000243485 Gene Expression GRCh38
FAM138A ENSG00000237613 Gene Expression GRCh38
OR4F5 ENSG00000186092 Gene Expression GRCh38
....
```
Best,
Kyle Hi @narsapvi,
I will have to defer your question over to @ktravaglini Hello @vchung,
Does the below error suggest that the below colns are not available in the test/validation set from leaderboard round..?!
ValueError: feature_names mismatch: ['prop_Class_Neuronal: GABAergic', 'prop_Class_Neuronal: Glutamatergic', 'prop_Class_Non-neuronal and Non-neural', 'prop_Subclass_Astrocyte', 'prop_Subclass_Chandelier', 'prop_Subclass_Endothelial',...
If we remove the above colns, our training accuracy will be effected.
Hi @narsapvi ,
In case it was missed in the other thread response, here are the logs:
9761247:
```
Traceback (most recent call last):
File "run_model.py", line 205, in
X, y, donor_id = preprocess_data(df)
File "run_model.py", line 192, in preprocess_data
known_cols = metadata_cols_present + excluded_cols + list(y.columns)
AttributeError: 'list' object has no attribute 'columns'
```
9761248:
```
File "/home/user/.local/lib/python3.8/site-packages/xgboost/core.py", line 3079, in _validate_features
raise ValueError(msg.format(self.feature_names, feature_names))
ValueError: feature_names mismatch: ['prop_Class_Neuronal: GABAergic', 'prop_Class_Neuronal: Glutamatergic', 'prop_Class_Non-neuronal and Non-neural', 'prop_Subclass_Astrocyte', 'prop_Subclass_Chandelier', 'prop_Subclass_Endothelial',...
```