Guide to install RAT's Dev Version

First create project directory structure.

mkdir rat3_project
cd rat3_project

mkdir models
mkdir data
mkdir params
mkdir secrets

We will use the rat3_project as the project directory for running RAT. We will use models to store the models vic, metsim, vic-routing and RAT. data, params and secrets directories would be used to store various sorts of data, parameter files and secrets repectively.

Next, we will clone RAT from github into the models directory.

git clone https://github.com/UW-SASWE/RAT.git models/RAT

Create the conda environment, and install RAT using conda develop.

conda create --prefix ./.env

conda activate ./.env
mamba env update --file models/RAT/environment.yml
mamba install conda-build
conda develop models/RAT/src/

Now you will be able to import rat and its components in the ./.env environment. Any changes you make in the RAT's source code in models/RAT/src/ directory will be imported. While installing rat as a conda package gives access to the cli command rat allowing interface to rat. The same behavior can be imitated in a development install by calling passing the arguments to the script models/RAT/src/rat/cli/rat_cli.py. Let's use this interface to initialize RAT. RAT requires some secrets to be able to use various functionalities, which have to be stored in the secrets/secrets.ini file. Read more about it .
```
conda activate ./.env
python models/RAT/src/rat/cli/rat_cli.py init -d . -g -s secrets/secrets.ini
```
RAT has not been initialized (hydrological models set up, global data downloaded and the parameter file has been populated). Follow instructions to complete the rest of the parameter file so that RAT can be run. To run RAT using the parameter file rat_config.yaml
```
conda activate ./.env
python models/RAT/src/rat/cli/rat_cli.py run -p params/rat_config.yaml
```