Skip to content

InspectOmop is a lightweight python 3 package that assists in the extraction of electronic health record(EHR) data from relational databases following the OHDSI OMOP Common Data Model(CDM) standard v>=5.

License

Notifications You must be signed in to change notification settings

jwimberl/InspectOMOP

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

InspectOMOP

Latest Release latest release
Package Status status
License license
Usage Stats usage

What is it?

inspectomop is a lightweight python 3 package that assists in the extraction of electronic health record(EHR) data from relational databases following the OHDSI OMOP Common Data Model(CDM) standard v>=5.

  • OHDSI: Observational Health Data Sciences and Informatics
  • OMOP: Observation Medical Outcomes Partnership

Why was this built?

A large portion of data science research is spent on ETL (Extraction, Transformation, and Loading). If the data are stored in a relational database, the first step includes deciphering the database schema and figuring out how to write SQL queries that will properly gather the information of interest. This can be both laborious and time consuming. inspectomop attempts to simplify extracting data from the OMOP CDM with an API that is easy to use, extensible, and SQL dialect agnostic.

One of the main benefits of adopting a CDM such as OMOP is that it promotes the sharing of ideas and methodology. Queries in inspectomop are simple python functions so using sqlAlchemy any user can create custom queries that can be shared across institutions and database management systems.

def my_query(inputs, inspector):

    # create SQL agnostic query usually of the form

    statement = select([columns]).where(inputs == criteria)

    return inspector.execute(statement)

Who is this for?

inspectomop is for any python 3 programmer with an interest in interfacing with an EHR relational database formatted to follow the OMOP CDM standard.

The OHDSI group has developed an excellent library of tools and methods written in R, but there are few, if any tools, for the python community.

Features

  • SQL dialect agnostic thanks to SQLAlchemy allowing for a variety of compatible database back ends
  • automatic reflection of DB tables to dot accessible python objects for easy traversal and inspection
  • preloaded with standard queries from the OHDSI group
  • results returnable as pandas dataframes or dataframe chunks for queries with a large number of rows
  • extensibility with custom queries built from simple python functions

SQL Dialect Compatibility

Below is a table comparing SQL dialect support for inspectomop versus the R SQLRender package written and maintained by the OHDSI group.

dialect inspectomop (python) SQLRender (R)
BigQuery No * Yes
Impala Yes * Yes
Netezza No * Yes
Oracle Yes Yes
PostgreSQL Yes Yes
Redshift Yes * Yes
SQL Server Yes Yes
SQLite Yes Unknown

Note: Compatibility is primarily based on the availability of dialects written for SQLAlchemy. Most have not bee explicitly tested by the author with the exception of SQLite v2.6.0 and SQL Server 2016 Service Pack 1 (13.0.4001.0). However, success stories and troubleshooting questions are welcome!

* BigQuery : python DB-API, but no sqlalchemy dialect as of 8/17/2018 (googleapis/google-cloud-python#3603)

* Impala : external dialect available via impyla package

* Netezza : python DB-API, but no sqlalchemy dialect as of 8/17/2018

* Redshift : external dialect available via sqlalchemy-redshift package

Where to get it

  • install from PyPI using pip with
pip install inspectomop

Dependencies

* Developed using SQLAlchemy 1.2.1 and Pandas 0.22.0

Documentation

Read the official documentation hosted on readthedocs for more information on usage and examples.

License

Feel free to fork, copy, share and contribute. This software released under GNU Affero GPL v3.0

About

InspectOmop is a lightweight python 3 package that assists in the extraction of electronic health record(EHR) data from relational databases following the OHDSI OMOP Common Data Model(CDM) standard v>=5.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%