new cohort sql file for v5.3.1. support and a cdm version field in the config file

[egillax]

Hi everyone,

My name is Egill, a postdoc from Erasmus MC in Rotterdam and I was testing your package against some synthetic data. I will eventually also test it on real data.

I needed to make some slight changes to the sql code so it would run on v5.3.1 of the cdm. I think the most common version in use is v5.3.1 so it would be good to support that. The only relevant difference from v5.3.1 to v6 is that the death_dates are not in the person table but in a death table. So I added a new cohort sql file to support that. Then I added a cdm version field in the config.

Then when running the notebooks I just added something like:

if config.CDM_VERSION == 'v5.3.1': 
    cohort_script_path = config.SQL_PATH_COHORTS + '/gen_EOL_cohort_v531.sql'  
else:  
    cohort_script_path = config.SQL_PATH_COHORTS + '/gen_EOL_cohort.sql'  

I didn't include the notebooks in the pull request since then I would overwrite the figures which are there, and since the data I used was synthetic my figures were not so good to have as an example.

Cheers,
Egill

[justinlimkz]

Hi Egill, thank you very much for testing out the package and adding v5.3.1 compatibility! I'm merging your pull request, and will edit the notebook to include the config.CDM_VERSION check you mentioned.