I am currently implementing the json feature importance output. It's unclear where the "concepts" table is expected to be.
Where can we find the concept table in the docker image? In the synthetic data, there is no concept.csv, instead there is a 'data_dictionary.csv' outside the main .tar which has different header fields than the expected concept table described here (https://www.synapse.org/#!Synapse:syn21849255/wiki/607301), specifically "code_name" in concept.csv vs. "concept_name" in data_dictionary.csv
Thanks
Created by Ivan Brugere ivanbrugere Hi @ivanbrugere,
I see, I don't think I understood what you were originally asking. Yes, you can use the `concept_name` field in the `data_dictionary.csv` or `concept.csv` files to fill out the `code_name` field in the `features.json` file. That was a mistake on my end. For the next iteration of this, we'll change it to `concept_name`.
Thank you,
@trberg I'm still not exactly sure about this. I downloaded "a" concept.csv file from Athena. The column in this file is "column_name" as in the data_dictionary.csv. Sorry to be picky about this, is the "code_name" column referenced in the export doc (https://www.synapse.org/#!Synapse:syn21849255/wiki/607301) incorrect or expecting something other than concept_name? Hi @ivanbrugere, I think you can always download a copy from the OHDSI web site (see https://forums.ohdsi.org/t/where-to-get-standard-vocabulary-files-such-as-domain-csv-and-concept-csv/3489 ).
Hi @trberg , I may have overlooked something. The concept.csv description doesn't seem to be in the documentation (https://www.synapse.org/#!Synapse:syn21849255/wiki/602415) or in the synthetic dataset (https://www.synapse.org/#!Synapse:syn22326807). Am I looking at the wrong pages?
Thanks,
Hi @ivanbrugere,
You can find the concept table in any of the `data` folders. They are located in the same directory as the other clinical tables. So you'll find `concept.csv` in those folders.
Thanks!
@trberg