Lung cancer submission error

Dear anajimenez

Thank you for sending me the error message. I have fix the problem and submit another one in Jan. 9, 2024, 8:48 a.m. Unfortunately, there are still something wrong in this algorithm.

Could you please help me find my log this time?

Thank you! wan77pc

Dear anajimenez,

Could you please help me with the error The algorithm failed on one or more cases., This submission was submited at Jan. 9, 2024, 4:50 p.m.

Thanks

Yang.

Dear @agaldran,

This is the trace raised by your container:

2024-01-08T17:52:54.886000+00:00 Traceback (most recent call last): 2024-01-08T17:52:54.886000+00:00 File "/opt/conda/lib/python3.10/runpy.py", line 196, in _run_module_as_main 2024-01-08T17:52:54.886000+00:00 return _run_code(code, main_globals, None, 2024-01-08T17:52:54.886000+00:00 File "/opt/conda/lib/python3.10/runpy.py", line 86, in _run_code 2024-01-08T17:52:54.886000+00:00 exec(code, run_globals) 2024-01-08T17:52:54.886000+00:00 File "/opt/app/process.py", line 243, in 2024-01-08T17:52:54.886000+00:00 Lungcancerosprediction().predict() 2024-01-08T17:52:54.886000+00:00 File "/opt/app/process.py", line 115, in predict 2024-01-08T17:52:54.886000+00:00 clinical_category = self.clinical_category_dict[clinical_category] 2024-01-08T17:52:54.886000+00:00 KeyError: None

It seems the "clinical_category" variable you are using to get the clinical info has a None value.

Hope this helps!

Dear @sss3barry,

I cannot see any log in your submission and the status of the analysis is "Cancelled" instead of "Failed". Can you try executing it again to see if it was a problem of the Gran Challenge platform?

Dear @ana,

Could you please also share my error log? I have tried submitting multiple times but I get the same error: "Failed: The algorithm failed on one or more cases". My docker file with local images works fine. I can't really debug this without the error log.thanks in advance

Hi Ana,

Thank you very much for sharing the log, it must be very inconvenient having to deal with all of us getting it wrong.

Because you told me I was getting a key error KeyError: None, I fixed my code by writing for this and all other variables the following:

try: clinical_category = self.clinical_category_dict[clinical_category]
except: clinical_category = 0

I have resubmitted but I still get the same uniformative "The algorithm failed on one or more cases" message. Maybe now I have something else wrong; can you please have a look? My submission was made on Jan. 9, 2024, 3:56 p.m.

Mil gracias!

Dear @chippy, It seems your are trying to store your inference results in the incorrect JSON file. Your submission is throwing the following error:

2024-01-09T08:50:31.877000+00:00 Traceback (most recent call last): 2024-01-09T08:50:31.877000+00:00 File "/usr/local/lib/python3.9/runpy.py", line 197, in _run_module_as_main 2024-01-09T08:50:31.877000+00:00 return _run_code(code, main_globals, None, 2024-01-09T08:50:31.877000+00:00 File "/usr/local/lib/python3.9/runpy.py", line 87, in _run_code 2024-01-09T08:50:31.877000+00:00 exec(code, run_globals) 2024-01-09T08:50:31.877000+00:00 File "/opt/app/process.py", line 193, in 2024-01-09T08:50:31.877000+00:00 Lungcancerosprediction().predict() 2024-01-09T08:50:31.877000+00:00 File "/opt/app/process.py", line 188, in predict 2024-01-09T08:50:31.877000+00:00 with open(predictions_json_path, 'w') as f: 2024-01-09T08:50:31.877000+00:00 PermissionError: [Errno 13] Permission denied: '/input/predictions.json'

Remember that the output needs to be stored in the file within the path '/output/overall-survival-months.json'

Dear @agaldran,

It seems your are having problems when importing SimpleITK library. Make sure you have it properly installed in your container.

This is the trace of your error:

2024-01-09T15:01:27.439000+00:00 Traceback (most recent call last): 2024-01-09T15:01:27.439000+00:00 File "/opt/conda/lib/python3.10/runpy.py", line 196, in _run_module_as_main 2024-01-09T15:01:27.439000+00:00 return _run_code(code, main_globals, None, 2024-01-09T15:01:27.439000+00:00 File "/opt/conda/lib/python3.10/runpy.py", line 86, in _run_code 2024-01-09T15:01:27.439000+00:00 exec(code, run_globals) 2024-01-09T15:01:27.439000+00:00 File "/opt/app/process.py", line 1, in 2024-01-09T15:01:27.439000+00:00 import SimpleITK as sitk 2024-01-09T15:01:27.439000+00:00 ModuleNotFoundError: No module named 'SimpleITK'

Thanks Ana, when rebuilding I had accidentally commented out the pip install requirements line from my DockerFile. My submission has now succeeded, thank you very much for your help!

Dear anajimenez,

I have resubmited to the lung validation track at Jan. 10, 2024, 8:52 a.m and failed again. The container ID is f869c39e-9b74-43c7-a04f-ccd74c549ef4. Could you please help me with the log?

My submission to Validation Prostate works well, that one only use mha file. Since some people report that at Prostate track the key age at train set is different to the patient_age at validation set, could you list all the available keys at validation and test set's json file? a sample file will be the best.

Best.

Dear @sss3barry, I see that in some cases you are outputting a negative survival time, so an invalid JSON file is created. Please, change your submission to force float positive values only.

With regards to prostate clinical data. Both validation and test phases include the following dictionary in the 'psa-and-age.json' file. Here is an example: {'psa': 82.86, 'patient_age': 67}

Hope this helps.

Dear anajimenez,

Thanks for your help, I have solved the issue and made a successful submission.

Best.

I submitted a new submission for validation lung cancer on Jan. 12, 2024, but I am receiving "Failed The algorithm failed on one or more cases.". Please help me find the logs!

Dear @ MaiAShaaban,

This is the trace of the issue causing your error:

2024-01-12T11:25:26.364000+00:00 Traceback (most recent call last): 2024-01-12T11:25:26.364000+00:00 File "/opt/conda/lib/python3.10/runpy.py", line 196, in _run_module_as_main 2024-01-12T11:25:26.364000+00:00 return _run_code(code, main_globals, None, 2024-01-12T11:25:26.364000+00:00 File "/opt/conda/lib/python3.10/runpy.py", line 86, in _run_code 2024-01-12T11:25:26.364000+00:00 exec(code, run_globals) 2024-01-12T11:25:26.364000+00:00 File "/opt/app/process.py", line 91, in 2024-01-12T11:25:26.364000+00:00 Lungcancerosprediction().predict() 2024-01-12T11:25:26.364000+00:00 File "/opt/app/process.py", line 54, in init 2024-01-12T11:25:26.364000+00:00 preprocess_csv = lung_mask_centerpoint(df_mask_stats) 2024-01-12T11:25:26.364000+00:00 File "/opt/app/preprocess.py", line 140, in lung_mask_centerpoint 2024-01-12T11:25:26.364000+00:00 x_mask_min_obj12 = np.min(df_mask_stats[["obj1_x_min", "obj2_x_min"]], axis=1) 2024-01-12T11:25:26.364000+00:00 File "/home/user/.local/lib/python3.10/site-packages/pandas/core/frame.py", line 3899, in getitem 2024-01-12T11:25:26.364000+00:00 indexer = self.columns._get_indexer_strict(key, "columns")[1] 2024-01-12T11:25:26.364000+00:00 File "/home/user/.local/lib/python3.10/site-packages/pandas/core/indexes/base.py", line 6115, in _get_indexer_strict 2024-01-12T11:25:26.364000+00:00 self._raise_if_missing(keyarr, indexer, axis_name) 2024-01-12T11:25:26.364000+00:00 File "/home/user/.local/lib/python3.10/site-packages/pandas/core/indexes/base.py", line 6176, in _raise_if_missing 2024-01-12T11:25:26.364000+00:00 raise KeyError(f"None of [{key}] are in the [{axis_name}]") 2024-01-12T11:25:26.364000+00:00 KeyError: "None of [Index(['obj1_x_min', 'obj2_x_min'], dtype='object')] are in the [columns]"

Dear Admins , Could you please share my error log related to the following submission : khatrid-lungsubmissionchamelion:releasev8 I made some significant changes to the code , which was earlier working , and is now failing on one or more cases. My local docker image works fine. Best, Dhruv