alumni_v2

This repo is used to deploy a reinforcement learning controller on the Alumni Hall. It will run following processes in parallel.

• Actual controller deployment script
• Online Data Collection script
• Offline Data collecting script
• Data-driven model learning script
• Controller learning script
• Off-line controller learning script

Installation of packages

1. Create virtual environment using (Optional but recommended step: This will have a clean and separate installation procedure for python packages that will not mess with existing applications in the server where it will run)

   $ python3 -m venv alumni_v2
   $ source alumni_v2/bin/activate

   ("python3 -m venv alumni_v2" might generate an error/warning on some Linux systems and it means an additional prerequisite needs to be fulfilled. I don't remember the exact details of the error as it has been a long time but in case it arises please get back to me with the error log and I will try to send out the fix.)

2. Install all requirements

   $ pip install -r requirements.txt

   or depending on your system,

   $ pip3 install -r requirements.txt

3. Exit the environment(Only to be done in case step 1 has been followed):

   $ deactivate

Starting the scripts

1. In case, virtual environment "alumni_v2" from step 1 has been created, you have to activate it at the location where it was created

   $ source alumni_v2/bin/activate

2. Launching online learning script(Mandatory)

   $ python online_learning.py

3. Start the script which calculates the wet bulb temperature in a seperate shell(Mandatory)

   $ python wbt_calculator.py

Step 4 is not needed for Alumni Deployment

4. In case you want to run the production facing server to visualize a live Dashboard, open a new terminal and execute the following

   $ waitress-serve --host <server ip address> --port <port to run> live_plot:app.server

   You can visualize the live dashboard by going to the IP address of the server followed by the port where you set it up.