Callbreak Tournament Starter Code

Starter codes for callbreak tournament

Contents

node_docker/ contains starter code for bot along with some instructions on building docker file

python_docker/ contains starter code for bot that utilizes sanic library along with some instructions.

sanic has been selected because it is considerably faster than flask but the downside is that the image size is now considerably bigger 60 MB -> 275 MB but runs ~9x faster on basic requests.

parsers/ contains script for parsing game history data . Use the parser to convert the game history data to sandbox/game api format.

Running instructions

Python

You will need to have python3 and pip3 installed to run this.

1. Navigate into python_docker first with cd python_docker/
2. Then, install requirements with pip install -r requirements.txt
3. Run with python src/app.py

JavaScript (Node)

You will need to have node >= 14 to run this.

1. Navigate into node_docker first with cd node_docker/
2. Run with node src/index.js

Parser

1. Copy the data from game analytics and save it in a file.

2. Usage : node stateToApiParser file round hand turn

   file refers to the filepath/file where the data from game analytics is saved.
   round ranges from 1 to 5
   hand ranges from 0 to 13 (0 stands for bidding phase)
   turn ranges from 0 to 3 (turn is in the respect to the playerIds)
   (turn 0 means turn of bot-1 when playerIds=['bot-1','bot-2','bot-3','bot-4'])
   

Building Instructions

You can build the existing projects by running docker build . -t <tag> inside node_docker/ or python_docker/.

. denotes that it will try to build the current directory

This builds you a docker image with the specified <tag> which you can view by running docker images.

You can use any name for <tag>. We reference the images using their id so, even if you submit your image with same tag name/ file name, it should not matter.

To run a built container run:

docker run --name <container name> -p 7000:7000 -m=2048m --cpus=1 -it <tag or image id>

This runs the image with <tag or image id> and binds local port 7000 with contianer port 7000, limits memory to 2040 MB RAM and 1 CPU. We run your docker image similarly on our server.

You should test your image by running it in the sandbox prior to uploading it.

After running your image, you create a container. The container reserves the name that you have provided it.

• To remove a container, run docker rm <container id or name>.
• To list all active containers, run dokcer ps.
• To list all containers, including inactive ones, run docker ps -a.

To export your docker image for submission, run docker save <tag> | gzip > <tag>.tar.gz.

Always save the docker file as .gz to reduce upload time.

To delete your existing docker image run docker image remove <image id>. You can get image id by running docker images