Error when running preprocessing

[vieanderes]

I am trying to synthesise the MIMIC IV data with the help of your code and README instructions.

After downloading the mimic dataset to the data/mimic4/raw location and setting the path environment variables I run:

python run_preprocess.py --dataset mimic4 --train_num 6000 --sample_num 10000

And get the following error:

FileNotFoundError: [Errno 2] No such file or directory: 'data/mimic4/raw/icd10-icd9.csv'

My directory structure is as follows:

Where is the icd10-icd9.csv file? I cannot find it in the mimic4 dataset.

Thank you already!

[LuChang-CS]

Hi, vandreslime, this icd10-icd9.csv is an unofficial mapping from ICD10 codes to ICD9 codes. I found it on the Internet, so I did not upload it to avoid some disagreement.

Here I uploaded it for now. You can download it for your project. In the future, I guess I will try to directly generate ICD10 codes for mimic4 without mapping them to ICD9.

[vieanderes]

Hi @LuChang-CS ,
great and thank you so much for the very quick response!

I got a new error:

File "/.../MTGAN/preprocess/parse_csv.py", line 201, in _load_icd_map
    icd_csv = pandas.read_csv(os.path.join(self.path, filename), usecols=cols, converters=converters)
...
ValueError: Usecols do not match columns, columns expected but not found: ['ICD9']

I guessed the last column name "Pure Victorian Logical" needed to be renamed to ICD9. Then I could successfully run the preprocessing.