id: 12456    nodeId: 12456    type: General    point: 357.0    linkPoint: .0    maker: cella    permission: linkable    made at: 2018.10.01 03:42    edited at: 2018.10.13 09:05
how to install EnergyPlus wrapper (rl-testbed-for-energyplus)
https://github.com/IBM/rl-testbed-for-energyplus
It is referenced from the paper of "Reinforcement Learning Testbed for Power-Consumption Optimization" by Moriyama et. al.
It is based on EnergyPlus 8.8.0
https://github.com/NREL/EnergyPlus/releases/tag/v8.8.0
During git cloning, a ssh key problem.
=> found a solution by associating a public key according to
https://stackoverflow.com/questions/2643502/git-permission-denied-publickey
"
First start by setting up your own public/private key pair set. This can use either DSA or RSA, so basically any key you setup will work. On most systems you can use ssh-keygen.

First you'll want to cd into your .ssh directory. Open up the terminal and run:
cd ~/.ssh && ssh-keygen
Next you need to copy this to your clipboard.
On OS X run: cat id_rsa.pub | pbcopy
On Linux run: cat id_rsa.pub | xclip
On Windows (via Cygwin/Git Bash) run: cat id_rsa.pub | clip
Add your key to your account via the website.
Finally setup your .gitconfig.
git config --global user.name "bob"
git config --global user.email bob@... (don't forget to restart your command line to make sure the config is reloaded)
Thats it you should be good to clone and checkout.
"
Just before source compilation, cmake needs to be installed.

Just before installing OpenAI gym baselines (cmake can be skipped as it is already installed as above, cf. the following git):
$ sudo apt install cmake libopenmpi-dev python3-dev zlib1g-dev
https://github.com/openai/baselines
Just before test run, a few modules and packages need to be installed:
$ pip3 install opencv-python
$ pip3 install pandas
$ pip3 install matplotlib
$ sudo apt install python3-tk

Return to Deep Reinforcement Learning: Pong from Pixels or how to install EnergyPlus wrapper (rl-testbed-for-energyplus)