The NEMS<->BAPHY API wraps separates the operation of BAPHY from that of NEMS so that neither system has to know about the internal workings of the other. This is good for both the modeler and data collector. As a NEMS modeler, we don't care how the data was captured -- we just want the data in a format that is friendly to NEMS. As maintainers of BAPHY, we don't care what analysis people do on the data -- we just don't want them accidentally writing garbage to our database.
SSH into the server that will be running this API.
Install the following required packages:
pip3 install flask --user
pip3 install flask_restful --user
Edit the configuration in "default_config.sh". Then run
source configs/default_config.sh # Sets environment variables
python3 -m nems_baphy_api
Note that NEMS_BAPHY_API_HOST should be localhost for testing, but be sure to make it an externally-facing address if you want anybody else to be able to use it.
In a new terminal on that same machine, you should be able to test that the API is working using:
curl http://localhost:3003/by-batch/273/gus018c-a3
where '273' is the batch and 'por39' is the cellid.
Alternatively, for static files,
curl http://localhost:3003/recording/TAR010c-18-1.tar.gz
If desired and time allows, it would not be that hard to add some "search" functionality to this that would help us find recordings based on:
-
stimulus type & cellid
-
cellid
-
area of brain