Skip to content

Living database of randomized controlled trials (RCTs)

Notifications You must be signed in to change notification settings

ijmarshall/trialstreamer

Repository files navigation

Trialstreamer

in progress...

To get started

  1. Install and run postgresql
  2. Create a database named 'trialstreamer'
  3. Make a copy of config.json.example to config.json, and populate
  4. Download RobotReviewer, and follow installation instructions (https://github.com/ijmarshall/robotreviewer)
  5. Run RobotReviewer in the REST API mode
  6. Run the following code to classify PubMed (typically 20 hours with GPU). This requires a proper setting of ictrp_retrieval_path, pubmed_local_data_path and pubmed_user_email configuration variables.
from trialstreamer import pubmed
pubmed.download_ftp_baseline()
  1. run python -m trialstreamer to start the server on localhost:5000

Running with Docker

Create a .env file from the .envTemplate file provided. Then, fill in the environment variables. NOTE: If a variable with the prefix TRIALSTREAMER_ was also defined in the config.json file, the environment variable will overwrite it. Make sure the pubmed_local_data_path configured at config.json exists and is the same as the one used as the volumes specified in docker compose files. It is recommended that this path is not located under trialstreamer folder.

To start the Trialstreamer API and Updates Crontab services, run the following commands after RobotReviewer is already running:

docker-compose build
docker-compose up -d 

or, as a single command:

docker-compose up --build -d 

If RobotReviewer is not running with docker-compose, you will have to create the network robotreviewer_default with:

docker network create robotreviewer_default 

Running with Docker in development mode

docker-compose -f docker-compose.dev.yml up --build -d

The development mode exposes the port 5000 and allows reloading the app when changes are detected, but the update script is not run automatically. However, the update script can be ran directly in the API docker container if necessary using the following command:

docker exec -ti trialstreamer_api_1 python update.py --source=<pubmed|medrxiv>

About

Living database of randomized controlled trials (RCTs)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •