havakv / havakv/pycox

AssertionError: assert durations.shape[0] == surv.shape[1] == surv_idx.shape[0] == events.shape[0]

Open
#159 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
995
Forks
203
PR merge metrics
No merged PRs in 30d

Description

[W 2023-04-23 10:28:38,810] Trial 2 failed with parameters: {'epochs': 1782, 'batch_norm': False, 'dropout': 0.3729027863986679, 'batch_size': 128, 'alpha': 0.20950098112214738, 'sigma': 0.021532851493345717, 'num_layers': 2, 'learning_rate': 0.0004455921289602272, 'optimizer_type': 'AdamWR', 'num_nodes_1': 56, 'num_nodes_2': 20} because of the following error: AssertionError().
Traceback (most recent call last):
File "[/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/optuna/study/_optimize.py](https://file+.vscode-resource.vscode-cdn.net/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/optuna/study/_optimize.py)", line 200, in _run_trial
value_or_values = func(trial)
File "[/tmp/ipykernel_25107/2073144205.py](https://file+.vscode-resource.vscode-cdn.net/tmp/ipykernel_25107/2073144205.py)", line 132, in objective
c_index = ev.concordance_td('antolini')
File "[/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/eval_surv.py](https://file+.vscode-resource.vscode-cdn.net/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/eval_surv.py)", line 191, in concordance_td
self._duration_idx(), method)
File "[/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/concordance.py](https://file+.vscode-resource.vscode-cdn.net/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/concordance.py)", line 93, in concordance_td
assert durations.shape[0] == surv.shape[1] == surv_idx.shape[0] == events.shape[0]
AssertionError
[W 2023-04-23 10:28:38,812] Trial 2 failed with value None.
[W 2023-04-23 10:28:46,183] Trial 1 failed with parameters: {'epochs': 1947, 'batch_norm': False, 'dropout': 0.29937082041280605, 'batch_size': 256, 'alpha': 0.3882041406677713, 'sigma': 0.7061223972910085, 'num_layers': 1, 'learning_rate': 0.001519199989009398, 'optimizer_type': 'AdamW', 'num_nodes_1': 54} because of the following error: AssertionError().
Traceback (most recent call last):
File "[/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/optuna/study/_optimize.py](https://file+.vscode-resource.vscode-cdn.net/home/huixian/anaconda3/envs/py3.7/lib/python3.7/site-packages/optuna/study/_optimize.py)", line 200, in _run_trial
value_or_values = func(trial)
File "[/tmp/ipykernel_25107/2073144205.py](https://file+.vscode-resource.vscode-cdn.net/tmp/ipykernel_25107/2073144205.py)", line 132, in objective
c_index = ev.concordance_td('antolini')
File "[/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/eval_surv.py](https://file+.vscode-resource.vscode-cdn.net/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/eval_surv.py)", line 191, in concordance_td
self._duration_idx(), method)
File "[/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/concordance.py](https://file+.vscode-resource.vscode-cdn.net/home/anaconda3/envs/py3.7/lib/python3.7/site-packages/pycox/evaluation/concordance.py)", line 93, in concordance_td
assert durations.shape[0] == surv.shape[1] == surv_idx.shape[0] == events.shape[0]
AssertionError

Contributor guide

No contributing guide indexed for this repository

Research direction

Start with pycox/evaluation/eval_surv.py at concordance_td('antolini') and pycox/evaluation/concordance.py around the reported assertion. Reproduce the failure from the Optuna objective using the supplied durations, survival predictions, indexes, and events, then trace why their dimensions differ. Done means the mismatch is explained and the relevant evaluation behavior is covered by a regression test.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.